Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.
Model information and context
The maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.
The model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.
The supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.
Chat, vision and code tasks
The card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.
For a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.
The maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.
For a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.
Conversation preparation and training
The model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.
Maverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.
The release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.
Practical result design
A document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.
Define how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.
Specific boundaries
The documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.
The card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.
", "aboutText": "Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.\n\nModel information and context\n\nThe maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.\n\nThe model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.\n\nThe supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.\n\nChat, vision and code tasks\n\nThe card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.\n\nFor a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.\n\nThe maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.\n\nFor a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.\n\nConversation preparation and training\n\nThe model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.\n\nMaverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.\n\nThe release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.\n\nPractical result design\n\nA document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.\n\nDefine how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.\n\nSpecific boundaries\n\nThe documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.\n\nThe card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Explain a return policy in Spanish\n\nExplain this policy in Spanish to a customer, keeping product code ZX-14 unchanged: 'Unused ZX-14 items may be returned within 14 days of delivery. Personalized items are excluded. A purchase receipt is required.' Use three short bullets and retain the distinction between ordinary and personalized items. Base the answer only on this supplied policy.Three Spanish policy bullets retaining the code, time trigger, exclusion and receipt requirement.ZX-14 must stay exact, and the exception for personalized items must not disappear.llama-4-maverickLlama 4 Maverick is Meta's native text-and-image Instruct model with 128 experts, 17B active and 400B total parameters. Its maker documents a 1M context, multilingual text/code output, visual recognition, captioning and image questions in a distinct release from Scout.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.
Model information and context
The maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.
The model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.
The supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.
Chat, vision and code tasks
The card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.
For a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.
The maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.
For a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.
Conversation preparation and training
The model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.
Maverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.
The release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.
Practical result design
A document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.
Define how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.
Specific boundaries
The documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.
The card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits llama-4-maverick | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits Minimum items: 1 | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits Model-specific | Default— |
Fieldmessages[].content | Typestring | RequirementOptional | Values & limits Applies when: Text conversation | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
Text · UTF8 text
messages[].content
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages |
Parametermaterial_types | Contracttext |
Parametermessages | Contractmessage array with role/content shape |
Parameterfile_material | Contractactual UTF-8 document inserted as literal text; no file ID inferred |
Parameteroptional_controls | Contractstream |
Parametercontrol_scope | ContractText conversation; stream is a boolean and defaults to false. |
Parameteroutput_modalities | Contracttext |
Parameterhosted_operations | Contractfalse |
Parametermultimodal_input_binding | Contractnull |
Method aliases: /chat/completions
Lifecycle: synchronous_or_streamed
Send exact model ID with messages.
Context length is not an output cap.
A text file may supply literal text; image/audio/file-object inputs need exact binding.
Tariff references do not add request options. Supported parameters are listed for the selected method.
Match each charge to the conditions shown. Use one rate for each set of alternative configurations.
| Configuration | Rate schedule | Context tokens, more than | Context tokens, up to and including | Charge |
|---|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Context tokens, more than— | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
| ConfigurationRequest-length tier 1 | Rate schedule— | Context tokens, more than0 | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model":"llama-4-maverick","messages":[{"role":"user","content":"hello"}],"stream":false}')\n\nPath("request.json").write_text(json.dumps(p,ensure_ascii=False))\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\nr=json.loads(Path("result.json").read_text())\nfor x in r.get("choices",[]):\n if x.get("message",{}).get("content"): print(x["message"]["content"])\nPY_RESULT\nfrom openai import OpenAI\nimport os\n\nif not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")\nclient=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)\nresult=client.chat.completions.create(model="llama-4-maverick", messages=[{"role":"user","content":"hello"}], stream=False)\nimport base64, json\nfrom pathlib import Path\n\ndef save_result(data, stem="result"):\n # Keep every field, including usage, IDs, reasoning and tool requests.\n Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")\n print(json.dumps(data, ensure_ascii=False, indent=2))\n def save_blob(encoded, suffix, label):\n raw = base64.b64decode(encoded, altchars=b"-_", validate=True)\n ext = "png" if raw.startswith(b"\\x89PNG\\r\\n\\x1a\\n") else "jpg" if raw.startswith(b"\\xff\\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix\n file = Path(stem + "-" + label + "." + ext)\n file.write_bytes(raw)\n print("Saved", file)\n for index, block in enumerate(data.get("content") or []):\n print("Content", index, json.dumps(block, ensure_ascii=False))\n for choice in data.get("choices") or []:\n print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))\n for index, item in enumerate(data.get("output") or []):\n print("Output", index, json.dumps(item, ensure_ascii=False))\n if item.get("type") == "image_generation_call" and item.get("result"):\n save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))\n for ci, candidate in enumerate(data.get("candidates") or []):\n for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):\n print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))\n blob = part.get("inlineData") or part.get("inline_data")\n if blob and blob.get("data"):\n mime = blob.get("mimeType") or blob.get("mime_type")\n suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")\n save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))\n\nsave_result(result.to_dict(mode="json"))\njson or SSE
Keep every complete returned response, file and content part.
Keep every choice and its content/tool calls/reasoning. In SSE assemble deltas by choice/tool index, retain the final finish reason and usage.
Keep request/response IDs and headers. No generic text job retrieval URL is supplied.
| Field | Type |
|---|---|
Fieldchoices[].index | Typeinteger |
Fieldchoices[].message.content | Typestring|null |
Fieldchoices[].message.tool_calls | Typetool_call[] |
Fieldchoices[].message.reasoning_content | Typestring |
Fieldchoices[].message.reasoning_details | Typeobject|array |
Fieldchoices[].finish_reason | Typestring|null |
Fieldusage | Typeobject |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Ailoot Ultra llama-4-maverick, meta-llama/llama-4-maverick
/chat/completions
llama-4-maverickLlama 4 Maverick is Meta's native text-and-image Instruct model with 128 experts, 17B active and 400B total parameters. Its maker documents a 1M context, multilingual text/code output, visual recognition, captioning and image questions in a distinct release from Scout.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.
Model information and context
The maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.
The model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.
The supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.
Chat, vision and code tasks
The card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.
For a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.
The maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.
For a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.
Conversation preparation and training
The model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.
Maverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.
The release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.
Practical result design
A document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.
Define how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.
Specific boundaries
The documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.
The card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits llama-4-maverick | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits user, assistant | Default— |
Fieldmessages[].content | Typestring | RequirementRequired | Values & limits | Default— |
Fieldmax_tokens | Typeinteger | RequirementRequired | Values & limits | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
Text · UTF-8 text
messages[].content string
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages, max_tokens |
Parameteroptional_controls | Contractstream |
Parametermaterial_types | Contracttext |
Parametertext_file_material | ContractRead the actual UTF-8 document and put its text in the indicated input or message content string. |
Lifecycle: synchronous_or_streamed
Tariff references do not add request options. Supported parameters are listed for the selected method.
Match each charge to the conditions shown. Use one rate for each set of alternative configurations.
| Configuration | Rate schedule | Context tokens, more than | Context tokens, up to and including | Charge |
|---|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Context tokens, more than— | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
| ConfigurationRequest-length tier 1 | Rate schedule— | Context tokens, more than0 | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "llama-4-maverick",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 document}"\nexport AILOOT_TEXT_FILE\npython3 - <<'PY_REQUEST'\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'llama-4-maverick', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nPath('request.json').write_text(json.dumps(payload, ensure_ascii=False), encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/messages \\\n -H "x-api-key: $AILOOT_API_KEY" -H 'anthropic-version: 2023-06-01' -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nPY_RESULT\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'llama-4-maverick', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nfrom urllib.request import Request, urlopen\nfrom urllib.error import HTTPError\n\napi_key = os.environ['AILOOT_API_KEY']\nif not api_key.strip():\n raise SystemExit('Set AILOOT_API_KEY before making a request')\nheaders = {'x-api-key': 'PLACEHOLDER_KEY', 'anthropic-version': '2023-06-01', 'Content-Type': 'application/json'}\nheaders['x-api-key'] = api_key\nrequest = Request('https://api.ailoot.ai/v1/messages', data=json.dumps(payload, ensure_ascii=False).encode('utf-8'), headers=headers, method='POST')\ntry:\n with urlopen(request, timeout=120) as response:\n raw = response.read()\n Path('response-headers.txt').write_text(str(response.headers), encoding='utf-8')\nexcept HTTPError as error:\n Path('result.json').write_bytes(error.read())\n Path('response-headers.txt').write_text(str(error.headers), encoding='utf-8')\n raise\nPath('result.json').write_bytes(raw)\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nJSON or SSE
Keep every complete returned response, file and content part.
Keep all content blocks, thinking signatures and actual tool-use IDs. Assemble all SSE block deltas; a first text block is not the complete output.
Keep request/response IDs and headers. No generic text job retrieval URL is supplied.
| Field | Type |
|---|---|
Fieldid | Typestring |
Fieldmodel | Typestring |
Fieldcontent[] | Typetext|thinking|tool_use|other content block |
Fieldstop_reason | Typestring|null |
Fieldusage | Typeobject |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Ailoot Ultra llama-4-maverick, meta-llama/llama-4-maverick
/chat/completions
meta-llama/llama-4-maverickLlama 4 Maverick is Meta's native text-and-image Instruct model with 128 experts, 17B active and 400B total parameters. Its maker documents a 1M context, multilingual text/code output, visual recognition, captioning and image questions in a distinct release from Scout.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.
Model information and context
The maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.
The model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.
The supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.
Chat, vision and code tasks
The card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.
For a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.
The maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.
For a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.
Conversation preparation and training
The model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.
Maverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.
The release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.
Practical result design
A document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.
Define how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.
Specific boundaries
The documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.
The card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits meta-llama/llama-4-maverick | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits Minimum items: 1 | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits Examples: system, user, assistant, tool, function Conditions: These roles are explicitly represented by Source compatibility types; the exact model role set remains model-specific. Model-specific | Default— |
Fieldmessages[].content | Typestring|content_part[]|null | RequirementOptional | Values & limits | Default— |
Fieldmessages[].tool_calls | Typetool_call[] | RequirementOptional | Values & limits | Default— |
Fieldmessages[].tool_call_id | Typestring | RequirementOptional | Values & limits Applies when: role=tool; use actual returned ID | Default— |
Fieldmessages[].name | Typestring | RequirementOptional | Values & limits | Default— |
Fieldmessages[].reasoning_content | Typestring | RequirementOptional | Values & limits Applies when: Round-trip actual returned reasoning where this model uses it | Default— |
Fieldmessages[].reasoning_details | Typeobject|array | RequirementOptional | Values & limits Model-specific | Default— |
Fieldmax_tokens | Typeinteger | RequirementOptional | Values & limits Minimum: 1 Model-specific | Default— |
Fieldmax_completion_tokens | Typeinteger | RequirementOptional | Values & limits Minimum: 1 Model-specific | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
Fieldstream_options.include_usage | Typeboolean | RequirementOptional | Values & limits Model-specific | Default— |
Fieldtemperature | Typenumber | RequirementOptional | Values & limits Model-specific | Default— |
Fieldtop_p | Typenumber | RequirementOptional | Values & limits Model-specific | Default— |
Fieldstop | Typestring|string[] | RequirementOptional | Values & limits Model-specific | Default— |
Fieldtools | Typefunction_tool[] | RequirementOptional | Values & limits Model-specific | Default— |
Fieldtools[].type | Typestring | RequirementOptional | Values & limits function Model-specific | Default— |
Fieldtools[].function.name | Typestring | RequirementRequired | Values & limits Applies when: function tool supplied | Default— |
Fieldtools[].function.description | Typestring | RequirementOptional | Values & limits | Default— |
Fieldtools[].function.parameters | TypeJSON_schema | RequirementOptional | Values & limits | Default— |
Fieldtool_choice | Typestring|object | RequirementOptional | Values & limits Model-specific | Default— |
Fieldparallel_tool_calls | Typeboolean | RequirementOptional | Values & limits Model-specific | Default— |
Fieldresponse_format | Typeobject | RequirementOptional | Values & limits Model-specific | Default— |
Fieldreasoning_effort | Typestring | RequirementOptional | Values & limits Model-specific | Default— |
Fieldreasoning | Typeobject | RequirementOptional | Values & limits Model-specific | Default— |
Fieldservice_tier | Typestring | RequirementOptional | Values & limits Model-specific | Default— |
Fieldinstructions | Typestring | RequirementOptional | Values & limits Compatibility field | Default— |
Fieldfunctions | Typefunction[] | RequirementOptional | Values & limits Model-specific | Default— |
Fieldfunction_call | Typestring|object | RequirementOptional | Values & limits Model-specific | Default— |
Fieldprompt_cache_options | Typeobject | RequirementOptional | Values & limits Model-specific | Default— |
Fieldmodalities | Typestring[] | RequirementOptional | Values & limits text | Default— |
Fieldprovider | Typeobject | RequirementOptional | Values & limits Conditions: provider.options is rejected on this Ultra text route | Default— |
Fieldcache_control.ttl | Typestring | RequirementOptional | Values & limits Conditions: 1h is rejected in actual prompt-cache locations | Default— |
Text · UTF8 text
messages[].content
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages |
Parametermaterial_types | Contracttext |
Parametermessages | Contractmessage array with role/content shape |
Parameterfile_material | Contractactual UTF-8 document inserted as literal text; no file ID inferred |
Parameteroptional_controls | Contractstream, temperature, top_p, stop, max_tokens, max_completion_tokens, reasoning, response_format, tools, tool_choice |
Parametercontrol_scope | Contractmodel-specific; user function tools differ from hosted tools |
Parameteroutput_modalities | Contracttext |
Parameterhosted_operations | Contractfalse |
Parametermultimodal_input_binding | Contractnull |
Lifecycle: synchronous_or_streamed
Send exact model ID with messages.
Context length is not an output cap.
A text file may supply literal text; image/audio/file-object inputs need exact binding.
Tariff references do not add request options. Supported parameters are listed for the selected method.
Match each charge to the conditions shown. Use one rate for each set of alternative configurations.
| Configuration | Rate schedule | Context tokens, more than | Context tokens, up to and including | Charge |
|---|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Context tokens, more than— | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
| ConfigurationRequest-length tier 1 | Rate schedule— | Context tokens, more than0 | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model": "meta-llama/llama-4-maverick", "messages": [{"role": "user", "content": "Explain the difference between an observation and a hypothesis in three sentences."}]}')\n\nPath("request.json").write_text(json.dumps(p,ensure_ascii=False))\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \\\n -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\nr=json.loads(Path("result.json").read_text())\nfor x in r.get("choices",[]):\n if x.get("message",{}).get("content"): print(x["message"]["content"])\nPY_RESULT\nfrom openai import OpenAI\nimport os\n\nif not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")\nclient=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)\nresult=client.chat.completions.create(model='meta-llama/llama-4-maverick',messages=[{"role":"user","content":"Explain observations and hypotheses in three sentences."}])\nimport base64, json\nfrom pathlib import Path\n\ndef save_result(data, stem="result"):\n # Keep every field, including usage, IDs, reasoning and tool requests.\n Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")\n print(json.dumps(data, ensure_ascii=False, indent=2))\n def save_blob(encoded, suffix, label):\n raw = base64.b64decode(encoded, altchars=b"-_", validate=True)\n ext = "png" if raw.startswith(b"\\x89PNG\\r\\n\\x1a\\n") else "jpg" if raw.startswith(b"\\xff\\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix\n file = Path(stem + "-" + label + "." + ext)\n file.write_bytes(raw)\n print("Saved", file)\n for index, block in enumerate(data.get("content") or []):\n print("Content", index, json.dumps(block, ensure_ascii=False))\n for choice in data.get("choices") or []:\n print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))\n for index, item in enumerate(data.get("output") or []):\n print("Output", index, json.dumps(item, ensure_ascii=False))\n if item.get("type") == "image_generation_call" and item.get("result"):\n save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))\n for ci, candidate in enumerate(data.get("candidates") or []):\n for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):\n print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))\n blob = part.get("inlineData") or part.get("inline_data")\n if blob and blob.get("data"):\n mime = blob.get("mimeType") or blob.get("mime_type")\n suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")\n save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))\n\nsave_result(result.to_dict(mode="json"))\nfrom openai import OpenAI\nimport os\nfrom pathlib import Path\n\nif not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")\nclient=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)\nimport base64, json\nfrom pathlib import Path\n\ndef save_result(data, stem="result"):\n # Keep every field, including usage, IDs, reasoning and tool requests.\n Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")\n print(json.dumps(data, ensure_ascii=False, indent=2))\n def save_blob(encoded, suffix, label):\n raw = base64.b64decode(encoded, altchars=b"-_", validate=True)\n ext = "png" if raw.startswith(b"\\x89PNG\\r\\n\\x1a\\n") else "jpg" if raw.startswith(b"\\xff\\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix\n file = Path(stem + "-" + label + "." + ext)\n file.write_bytes(raw)\n print("Saved", file)\n for index, block in enumerate(data.get("content") or []):\n print("Content", index, json.dumps(block, ensure_ascii=False))\n for choice in data.get("choices") or []:\n print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))\n for index, item in enumerate(data.get("output") or []):\n print("Output", index, json.dumps(item, ensure_ascii=False))\n if item.get("type") == "image_generation_call" and item.get("result"):\n save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))\n for ci, candidate in enumerate(data.get("candidates") or []):\n for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):\n print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))\n blob = part.get("inlineData") or part.get("inline_data")\n if blob and blob.get("data"):\n mime = blob.get("mimeType") or blob.get("mime_type")\n suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")\n save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))\n\nwith client.chat.completions.stream(model='meta-llama/llama-4-maverick',messages=[{"role":"user","content":"Explain observations and hypotheses in three sentences."}]) as stream:\n with Path("stream-events.jsonl").open("w",encoding="utf-8") as log:\n for event in stream:\n payload=event.to_dict(mode="json")\n log.write(json.dumps(payload,ensure_ascii=False)+"\\n"); log.flush()\n print(json.dumps(payload,ensure_ascii=False))\n final=stream.get_final_completion()\n save_result(final.to_dict(mode="json"),"stream-result")\njson or SSE
Keep every complete returned response, file and content part.
Keep every choice and its content/tool calls/reasoning. In SSE assemble deltas by choice/tool index, retain the final finish reason and usage.
Keep request/response IDs and headers. No generic text job retrieval URL is supplied.
| Field | Type |
|---|---|
Fieldchoices[].index | Typeinteger |
Fieldchoices[].message.content | Typestring|null |
Fieldchoices[].message.tool_calls | Typetool_call[] |
Fieldchoices[].message.reasoning_content | Typestring |
Fieldchoices[].message.reasoning_details | Typeobject|array |
Fieldchoices[].finish_reason | Typestring|null |
Fieldusage | Typeobject |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Ailoot Ultra llama-4-maverick, meta-llama/llama-4-maverick
/chat/completions
meta-llama/llama-4-maverickLlama 4 Maverick is Meta's native text-and-image Instruct model with 128 experts, 17B active and 400B total parameters. Its maker documents a 1M context, multilingual text/code output, visual recognition, captioning and image questions in a distinct release from Scout.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Llama 4 Maverick Instruct is Meta's natively multimodal model for assistant-style conversation and visual reasoning. It receives multilingual text and images and generates multilingual text or code. Its mixture-of-experts architecture uses early fusion, combining the visual and textual information within the model rather than treating the image as an unrelated caption supplied by a separate text assistant.
Model information and context
The maker lists 128 experts, 17 billion active parameters and 400 billion total parameters for Maverick. These are distinct from Scout's 16-expert, 109B-total configuration even though both carry the 17B active-size label. The Instruct name identifies the assistant-oriented checkpoint rather than the pretrained Maverick model.
The model card documents a 1M context window for Maverick. Use the context to retain the relevant passages, images, conversation and requested answer together. A large source collection still needs a defined question and artifact. Label the source documents and images and request the location supporting each extracted finding instead of asking for an undirected summary of all material.
The supported languages are Arabic, English, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. The maker distinguishes these twelve supported languages from the broader collection of about two hundred languages used in pretraining. Training-language presence should not be turned into an equal capability promise for every language.
Chat, vision and code tasks
The card's intended visual uses include recognition, image reasoning, captioning and answering general questions about an image. For a photograph, ask the model to identify the objects or relationship relevant to the task. For a chart, request a comparison between named series, with the axes and units retained. For a diagram, ask which labeled connection supports a supplied requirement.
For a multi-image comparison, identify the role of each image and ask for the same criteria across them. For example, supply an original design and a revision, then request visible matches, differences and details the images do not establish. Keep the observations separate from conclusions that depend on the accompanying text. This makes the output easier to inspect against the actual source.
The maker reports testing image understanding with up to five input images. That is the documented tested scope, not an unconditional hard image-count maximum. Preserve the detail needed for the question and the order of the images. Adding a large number of unrelated images does not itself define what should be compared.
For a coding task, provide the requirement, relevant source and expected behavior. Ask for the proposed implementation separately from its assumptions and examples. A screenshot can supply visual requirements, while the textual condition supplies behavior that the screenshot cannot show. Generated code is a text artifact; running it, inspecting a rendered interface or applying it to a repository requires the connected environment.
Conversation preparation and training
The model card demonstrates the matching processor and chat template with two images and a written comparison question. Keep conversation roles and image-text association explicit. A system instruction can define the audience, tone and response structure, while the user message supplies the actual task and evidence. The maker describes post-training improvements to system-prompt steerability and conversational formatting.
Maverick is pretrained on approximately 22 trillion tokens of multimodal data, with a stated pretraining-data cutoff of August 2024. This is a static training-data boundary rather than live access to current information. When a question depends on newer or private material, provide that material as source context or through a separately connected retrieval operation.
The release includes BF16 and separately quantized FP8 weights. Those precision choices do not create another input modality or grant the 10M context of Scout to Maverick. Preserve the exact Instruct checkpoint and its configuration when comparing a repeated task or a derived deployment.
Practical result design
A document-image task can request the source identifier, visible statement, extracted value and interpretation as separate fields. A conversational explanation can retain the original terms and adapt the prose to the reader's supported language. A coding request can return the revised fragment and examples that distinguish its intended behavior.
Define how the final result will be used before supplying the material. A short caption, a comparison table and a proposed program solve different tasks even where they share the same images. The model's visual and language capabilities supply the interpretation and text; external retrieval, execution or file production comes from application components.
Specific boundaries
The documented inputs here are text and images, with text and code output. Audio or video understanding, speech generation and Muse Spark's named reasoning-effort controls should not be assigned to this Maverick edition. Llama Guard is also a separately trained classifier rather than an automatic classifier output from a Maverick answer.
The card's 256-token generation example is an example setting, and the sources read do not establish a separate hard output-token maximum beyond the model's context specification. Keep the Maverick identity, supported-language scope and source-linked final artifact explicit rather than borrowing limits or modes from another Llama release.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits meta-llama/llama-4-maverick | Default— |
Fieldmessages | Typemessage[] | RequirementRequired | Values & limits | Default— |
Fieldmessages[].role | Typestring | RequirementRequired | Values & limits user, assistant | Default— |
Fieldmessages[].content | Typestring | RequirementRequired | Values & limits | Default— |
Fieldmax_tokens | Typeinteger | RequirementRequired | Values & limits | Default— |
Fieldstream | Typeboolean | RequirementOptional | Values & limits | Defaultfalse |
Text · UTF-8 text
messages[].content string
| Parameter | Contract |
|---|---|
Parametercontent_type | Contractapplication/json |
Parameterrequired | Contractmodel, messages, max_tokens |
Parameteroptional_controls | Contractstream |
Parametermaterial_types | Contracttext |
Parametertext_file_material | ContractRead the actual UTF-8 document and put its text in the indicated input or message content string. |
Lifecycle: synchronous_or_streamed
Tariff references do not add request options. Supported parameters are listed for the selected method.
Match each charge to the conditions shown. Use one rate for each set of alternative configurations.
| Configuration | Rate schedule | Context tokens, more than | Context tokens, up to and including | Charge |
|---|---|---|---|---|
| ConfigurationBase token rates | Rate scheduleBase rate | Context tokens, more than— | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
| ConfigurationRequest-length tier 1 | Rate schedule— | Context tokens, more than0 | Context tokens, up to and includingNo upper bound | Charge Input · Input tokens $0.000375 / 1K tokens Output · Output tokens $0.001305 / 1K tokens Input · Cache read $0.0001 / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "meta-llama/llama-4-maverick",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\n: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 document}"\nexport AILOOT_TEXT_FILE\npython3 - <<'PY_REQUEST'\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'meta-llama/llama-4-maverick', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nPath('request.json').write_text(json.dumps(payload, ensure_ascii=False), encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/messages \\\n -H "x-api-key: $AILOOT_API_KEY" -H 'anthropic-version: 2023-06-01' -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nPY_RESULT\nimport json, os\nfrom pathlib import Path\n\nsource_file = Path(os.environ['AILOOT_TEXT_FILE'])\nif not source_file.is_file():\n raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')\ntext = source_file.read_text(encoding='utf-8')\nif not text.strip():\n raise SystemExit('Choose a document containing text')\npayload = {'model': 'meta-llama/llama-4-maverick', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}\npayload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\\n\\n' + text\nfrom urllib.request import Request, urlopen\nfrom urllib.error import HTTPError\n\napi_key = os.environ['AILOOT_API_KEY']\nif not api_key.strip():\n raise SystemExit('Set AILOOT_API_KEY before making a request')\nheaders = {'x-api-key': 'PLACEHOLDER_KEY', 'anthropic-version': '2023-06-01', 'Content-Type': 'application/json'}\nheaders['x-api-key'] = api_key\nrequest = Request('https://api.ailoot.ai/v1/messages', data=json.dumps(payload, ensure_ascii=False).encode('utf-8'), headers=headers, method='POST')\ntry:\n with urlopen(request, timeout=120) as response:\n raw = response.read()\n Path('response-headers.txt').write_text(str(response.headers), encoding='utf-8')\nexcept HTTPError as error:\n Path('result.json').write_bytes(error.read())\n Path('response-headers.txt').write_text(str(error.headers), encoding='utf-8')\n raise\nPath('result.json').write_bytes(raw)\nimport json\nfrom pathlib import Path\n\nresult = json.loads(Path('result.json').read_text(encoding='utf-8'))\nprint(json.dumps(result, ensure_ascii=False, indent=2))\nJSON or SSE
Keep every complete returned response, file and content part.
Keep all content blocks, thinking signatures and actual tool-use IDs. Assemble all SSE block deltas; a first text block is not the complete output.
Keep request/response IDs and headers. No generic text job retrieval URL is supplied.
| Field | Type |
|---|---|
Fieldid | Typestring |
Fieldmodel | Typestring |
Fieldcontent[] | Typetext|thinking|tool_use|other content block |
Fieldstop_reason | Typestring|null |
Fieldusage | Typeobject |
| Response | Condition | Next step |
|---|---|---|
Response400 · invalid_request_error | ConditionInvalid required fields, material representation or parameter value | Next stepCorrect the indicated fields before a new creation request. |
Ailoot Ultra llama-4-maverick, meta-llama/llama-4-maverick
/chat/completions
Technical example in English
Explain a return policy in Spanish\n\nExplain this policy in Spanish to a customer, keeping product code ZX-14 unchanged: 'Unused ZX-14 items may be returned within 14 days of delivery. Personalized items are excluded. A purchase receipt is required.' Use three short bullets and retain the distinction between ordinary and personalized items. Base the answer only on this supplied policy.Three Spanish policy bullets retaining the code, time trigger, exclusion and receipt requirement.ZX-14 must stay exact, and the exception for personalized items must not disappear.