Ministral 3 8B Instruct 2512 is the middle-sized instruction-tuned model in Mistral AI's Ministral 3 family. The released checkpoint is post-trained for chat and instruction tasks in FP8 precision. It combines a language component with a vision encoder, accepting text and images and returning text rather than an image or audio artifact.
Architecture and key features
The maker lists an 8.4-billion-parameter language model and a 0.4-billion-parameter vision encoder. Its context window is described as 256k, and the published usage notes use a default configured length of 262,144. These describe this 8B-size edition; the smaller 3B and larger 14B family members retain their own component sizes and configurations.
Vision extends the model's input beyond a written conversation. A document image can be paired with a question about a selected field, paragraph or visible relationship. A screenshot can be paired with the user's goal and a request to identify the relevant control. Ask for the observation supporting each conclusion when the answer depends on visual detail, and preserve headings, labels and units needed to interpret a crop.
The card describes multilingual work across dozens of languages, including English, French, Spanish, German, Italian, Portuguese, Dutch, Chinese, Japanese, Korean and Arabic. It also identifies adherence to system instructions, native function calling and JSON output. Define the final language, role and output requirements explicitly rather than leaving the model to infer the intended artifact from a large input.
Document, conversation and agent tasks
The maker's use cases include chat interfaces, everyday assistants, image and document description, translation, content generation and specialized agents. For a document assistant, provide the page and specify the information to return. An extraction request can name the field labels, expected types and treatment of missing values. A useful answer might contain the document identifier, the extracted value and a short explanation tied to the visible source.
For an ongoing conversation, keep the relevant prior answer and ask a focused follow-up. A user can first request a document summary and then ask which statement supports one conclusion. This changes the required output from a general overview to a source-linked explanation while retaining the original material and the purpose of the dialogue.
For multilingual editing, supply the original passage, the intended reader and a terminology list. Request a revised version that retains names, conditions and units, followed by a brief explanation of the changes if useful. For content generation, specify the audience and the facts that should appear before asking for a short draft or alternative wording.
In an agent task, the application provides the tools and their parameters. The model can choose a function and prepare arguments, while the application executes it and returns the result. The card's function example combines an image question with a calculation tool. Keep the tool result separate from the proposed call so that the subsequent written answer uses the information actually returned by the operation.
Recommended preparation and settings
The maker recommends a system instruction that describes the environment, use case and how tools should be used. Keep the set of tools well-defined and limited to the minimum needed for the task. A small lookup workflow can offer the functions relevant to the current question, with clearly named required arguments, instead of presenting a large collection of unrelated operations.
For everyday and production-style use, the card recommends a temperature below 0.1, with higher temperatures available for creative exploration. Treat this as a recommendation for the instruction workflow, not a separate reasoning-mode control. The family has distinct Base, Instruct and Reasoning checkpoints; this entry is the Instruct 2512 edition, not the separate Reasoning release of the same size.
For images, the maker recommends aspect ratios near 1:1 and avoiding very thin or very wide inputs. Crop when needed while retaining the labels and context required by the question. If a page contains several unrelated areas, identify the part to inspect and the result to return. A legible, focused image is more useful than asking the model to guess which small detail in a broad page matters.
Context and edition boundaries
Plan the text, images, conversation and desired answer around the configured context. The usage notes explicitly allow reducing the configured model length when less context is needed, so the full window is not an instruction to fill every request with all available material. The card's 512-token generation example is only an example setting, and the sources read do not establish a separate hard maximum answer length.
The FP8 Instruct checkpoint, the pretrained Base release and the Reasoning release are separate editions. The maker also describes specialization and fine-tuning as use cases. Preserve the named checkpoint and its tokenizer when preparing the conversation, and keep text generation distinct from repository access, function execution or other actions supplied by the surrounding application.
", "aboutText": "Ministral 3 8B Instruct 2512 is the middle-sized instruction-tuned model in Mistral AI's Ministral 3 family. The released checkpoint is post-trained for chat and instruction tasks in FP8 precision. It combines a language component with a vision encoder, accepting text and images and returning text rather than an image or audio artifact.\n\nArchitecture and key features\n\nThe maker lists an 8.4-billion-parameter language model and a 0.4-billion-parameter vision encoder. Its context window is described as 256k, and the published usage notes use a default configured length of 262,144. These describe this 8B-size edition; the smaller 3B and larger 14B family members retain their own component sizes and configurations.\n\nVision extends the model's input beyond a written conversation. A document image can be paired with a question about a selected field, paragraph or visible relationship. A screenshot can be paired with the user's goal and a request to identify the relevant control. Ask for the observation supporting each conclusion when the answer depends on visual detail, and preserve headings, labels and units needed to interpret a crop.\n\nThe card describes multilingual work across dozens of languages, including English, French, Spanish, German, Italian, Portuguese, Dutch, Chinese, Japanese, Korean and Arabic. It also identifies adherence to system instructions, native function calling and JSON output. Define the final language, role and output requirements explicitly rather than leaving the model to infer the intended artifact from a large input.\n\nDocument, conversation and agent tasks\n\nThe maker's use cases include chat interfaces, everyday assistants, image and document description, translation, content generation and specialized agents. For a document assistant, provide the page and specify the information to return. An extraction request can name the field labels, expected types and treatment of missing values. A useful answer might contain the document identifier, the extracted value and a short explanation tied to the visible source.\n\nFor an ongoing conversation, keep the relevant prior answer and ask a focused follow-up. A user can first request a document summary and then ask which statement supports one conclusion. This changes the required output from a general overview to a source-linked explanation while retaining the original material and the purpose of the dialogue.\n\nFor multilingual editing, supply the original passage, the intended reader and a terminology list. Request a revised version that retains names, conditions and units, followed by a brief explanation of the changes if useful. For content generation, specify the audience and the facts that should appear before asking for a short draft or alternative wording.\n\nIn an agent task, the application provides the tools and their parameters. The model can choose a function and prepare arguments, while the application executes it and returns the result. The card's function example combines an image question with a calculation tool. Keep the tool result separate from the proposed call so that the subsequent written answer uses the information actually returned by the operation.\n\nRecommended preparation and settings\n\nThe maker recommends a system instruction that describes the environment, use case and how tools should be used. Keep the set of tools well-defined and limited to the minimum needed for the task. A small lookup workflow can offer the functions relevant to the current question, with clearly named required arguments, instead of presenting a large collection of unrelated operations.\n\nFor everyday and production-style use, the card recommends a temperature below 0.1, with higher temperatures available for creative exploration. Treat this as a recommendation for the instruction workflow, not a separate reasoning-mode control. The family has distinct Base, Instruct and Reasoning checkpoints; this entry is the Instruct 2512 edition, not the separate Reasoning release of the same size.\n\nFor images, the maker recommends aspect ratios near 1:1 and avoiding very thin or very wide inputs. Crop when needed while retaining the labels and context required by the question. If a page contains several unrelated areas, identify the part to inspect and the result to return. A legible, focused image is more useful than asking the model to guess which small detail in a broad page matters.\n\nContext and edition boundaries\n\nPlan the text, images, conversation and desired answer around the configured context. The usage notes explicitly allow reducing the configured model length when less context is needed, so the full window is not an instruction to fill every request with all available material. The card's 512-token generation example is only an example setting, and the sources read do not establish a separate hard maximum answer length.\n\nThe FP8 Instruct checkpoint, the pretrained Base release and the Reasoning release are separate editions. The maker also describes specialization and fine-tuning as use cases. Preserve the named checkpoint and its tokenizer when preparing the conversation, and keep text generation distinct from repository access, function execution or other actions supplied by the surrounding application.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Extract a compact invoice record\n\nExtract the invoice identifier, supplier, total and currency from this text: Invoice INV-731. Supplier: Cedar Works. Subtotal: 80 EUR. Tax: 16 EUR. Total: 96 EUR. Payment due: not stated. Return those fields plus payment_due; use null when absent and preserve the identifier exactly.A labelled record with invoice_id, supplier, total, currency and payment_due, using null for the unstated date.Check that the total is taken from the Total line rather than the subtotal and that EUR is retained.mistralai/ministral-8b-2512Ministral 3 8B Instruct 2512 is a multilingual text-and-image instruction model with an 8.4B language component and a 0.4B vision encoder. It supports a 256k maker context, system instructions, JSON and native function calling for assistants and document workflows.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Ministral 3 8B Instruct 2512 is the middle-sized instruction-tuned model in Mistral AI's Ministral 3 family. The released checkpoint is post-trained for chat and instruction tasks in FP8 precision. It combines a language component with a vision encoder, accepting text and images and returning text rather than an image or audio artifact.
Architecture and key features
The maker lists an 8.4-billion-parameter language model and a 0.4-billion-parameter vision encoder. Its context window is described as 256k, and the published usage notes use a default configured length of 262,144. These describe this 8B-size edition; the smaller 3B and larger 14B family members retain their own component sizes and configurations.
Vision extends the model's input beyond a written conversation. A document image can be paired with a question about a selected field, paragraph or visible relationship. A screenshot can be paired with the user's goal and a request to identify the relevant control. Ask for the observation supporting each conclusion when the answer depends on visual detail, and preserve headings, labels and units needed to interpret a crop.
The card describes multilingual work across dozens of languages, including English, French, Spanish, German, Italian, Portuguese, Dutch, Chinese, Japanese, Korean and Arabic. It also identifies adherence to system instructions, native function calling and JSON output. Define the final language, role and output requirements explicitly rather than leaving the model to infer the intended artifact from a large input.
Document, conversation and agent tasks
The maker's use cases include chat interfaces, everyday assistants, image and document description, translation, content generation and specialized agents. For a document assistant, provide the page and specify the information to return. An extraction request can name the field labels, expected types and treatment of missing values. A useful answer might contain the document identifier, the extracted value and a short explanation tied to the visible source.
For an ongoing conversation, keep the relevant prior answer and ask a focused follow-up. A user can first request a document summary and then ask which statement supports one conclusion. This changes the required output from a general overview to a source-linked explanation while retaining the original material and the purpose of the dialogue.
For multilingual editing, supply the original passage, the intended reader and a terminology list. Request a revised version that retains names, conditions and units, followed by a brief explanation of the changes if useful. For content generation, specify the audience and the facts that should appear before asking for a short draft or alternative wording.
In an agent task, the application provides the tools and their parameters. The model can choose a function and prepare arguments, while the application executes it and returns the result. The card's function example combines an image question with a calculation tool. Keep the tool result separate from the proposed call so that the subsequent written answer uses the information actually returned by the operation.
Recommended preparation and settings
The maker recommends a system instruction that describes the environment, use case and how tools should be used. Keep the set of tools well-defined and limited to the minimum needed for the task. A small lookup workflow can offer the functions relevant to the current question, with clearly named required arguments, instead of presenting a large collection of unrelated operations.
For everyday and production-style use, the card recommends a temperature below 0.1, with higher temperatures available for creative exploration. Treat this as a recommendation for the instruction workflow, not a separate reasoning-mode control. The family has distinct Base, Instruct and Reasoning checkpoints; this entry is the Instruct 2512 edition, not the separate Reasoning release of the same size.
For images, the maker recommends aspect ratios near 1:1 and avoiding very thin or very wide inputs. Crop when needed while retaining the labels and context required by the question. If a page contains several unrelated areas, identify the part to inspect and the result to return. A legible, focused image is more useful than asking the model to guess which small detail in a broad page matters.
Context and edition boundaries
Plan the text, images, conversation and desired answer around the configured context. The usage notes explicitly allow reducing the configured model length when less context is needed, so the full window is not an instruction to fill every request with all available material. The card's 512-token generation example is only an example setting, and the sources read do not establish a separate hard maximum answer length.
The FP8 Instruct checkpoint, the pretrained Base release and the Reasoning release are separate editions. The maker also describes specialization and fine-tuning as use cases. Preserve the named checkpoint and its tokenizer when preparing the conversation, and keep text generation distinct from repository access, function execution or other actions supplied by the surrounding application.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits mistralai/ministral-8b-2512 | 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 $0.00066 / 1K tokens Output · Output tokens $0.00066 / 1K tokens Input · Cache read $0.00003 / 1K tokens Input · Cache write $0 / 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 $0.00066 / 1K tokens Output · Output tokens $0.00066 / 1K tokens Input · Cache read $0.00003 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model": "mistralai/ministral-8b-2512", "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='mistralai/ministral-8b-2512',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='mistralai/ministral-8b-2512',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 mistralai/ministral-8b-2512
Use the method paths shown above.
mistralai/ministral-8b-2512Ministral 3 8B Instruct 2512 is a multilingual text-and-image instruction model with an 8.4B language component and a 0.4B vision encoder. It supports a 256k maker context, system instructions, JSON and native function calling for assistants and document workflows.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Ministral 3 8B Instruct 2512 is the middle-sized instruction-tuned model in Mistral AI's Ministral 3 family. The released checkpoint is post-trained for chat and instruction tasks in FP8 precision. It combines a language component with a vision encoder, accepting text and images and returning text rather than an image or audio artifact.
Architecture and key features
The maker lists an 8.4-billion-parameter language model and a 0.4-billion-parameter vision encoder. Its context window is described as 256k, and the published usage notes use a default configured length of 262,144. These describe this 8B-size edition; the smaller 3B and larger 14B family members retain their own component sizes and configurations.
Vision extends the model's input beyond a written conversation. A document image can be paired with a question about a selected field, paragraph or visible relationship. A screenshot can be paired with the user's goal and a request to identify the relevant control. Ask for the observation supporting each conclusion when the answer depends on visual detail, and preserve headings, labels and units needed to interpret a crop.
The card describes multilingual work across dozens of languages, including English, French, Spanish, German, Italian, Portuguese, Dutch, Chinese, Japanese, Korean and Arabic. It also identifies adherence to system instructions, native function calling and JSON output. Define the final language, role and output requirements explicitly rather than leaving the model to infer the intended artifact from a large input.
Document, conversation and agent tasks
The maker's use cases include chat interfaces, everyday assistants, image and document description, translation, content generation and specialized agents. For a document assistant, provide the page and specify the information to return. An extraction request can name the field labels, expected types and treatment of missing values. A useful answer might contain the document identifier, the extracted value and a short explanation tied to the visible source.
For an ongoing conversation, keep the relevant prior answer and ask a focused follow-up. A user can first request a document summary and then ask which statement supports one conclusion. This changes the required output from a general overview to a source-linked explanation while retaining the original material and the purpose of the dialogue.
For multilingual editing, supply the original passage, the intended reader and a terminology list. Request a revised version that retains names, conditions and units, followed by a brief explanation of the changes if useful. For content generation, specify the audience and the facts that should appear before asking for a short draft or alternative wording.
In an agent task, the application provides the tools and their parameters. The model can choose a function and prepare arguments, while the application executes it and returns the result. The card's function example combines an image question with a calculation tool. Keep the tool result separate from the proposed call so that the subsequent written answer uses the information actually returned by the operation.
Recommended preparation and settings
The maker recommends a system instruction that describes the environment, use case and how tools should be used. Keep the set of tools well-defined and limited to the minimum needed for the task. A small lookup workflow can offer the functions relevant to the current question, with clearly named required arguments, instead of presenting a large collection of unrelated operations.
For everyday and production-style use, the card recommends a temperature below 0.1, with higher temperatures available for creative exploration. Treat this as a recommendation for the instruction workflow, not a separate reasoning-mode control. The family has distinct Base, Instruct and Reasoning checkpoints; this entry is the Instruct 2512 edition, not the separate Reasoning release of the same size.
For images, the maker recommends aspect ratios near 1:1 and avoiding very thin or very wide inputs. Crop when needed while retaining the labels and context required by the question. If a page contains several unrelated areas, identify the part to inspect and the result to return. A legible, focused image is more useful than asking the model to guess which small detail in a broad page matters.
Context and edition boundaries
Plan the text, images, conversation and desired answer around the configured context. The usage notes explicitly allow reducing the configured model length when less context is needed, so the full window is not an instruction to fill every request with all available material. The card's 512-token generation example is only an example setting, and the sources read do not establish a separate hard maximum answer length.
The FP8 Instruct checkpoint, the pretrained Base release and the Reasoning release are separate editions. The maker also describes specialization and fine-tuning as use cases. Preserve the named checkpoint and its tokenizer when preparing the conversation, and keep text generation distinct from repository access, function execution or other actions supplied by the surrounding application.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits mistralai/ministral-8b-2512 | 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 $0.00066 / 1K tokens Output · Output tokens $0.00066 / 1K tokens Input · Cache read $0.00003 / 1K tokens Input · Cache write $0 / 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 $0.00066 / 1K tokens Output · Output tokens $0.00066 / 1K tokens Input · Cache read $0.00003 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "mistralai/ministral-8b-2512",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${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': 'mistralai/ministral-8b-2512', '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': 'mistralai/ministral-8b-2512', '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 mistralai/ministral-8b-2512
Use the method paths shown above.
Technical example in English
Extract a compact invoice record\n\nExtract the invoice identifier, supplier, total and currency from this text: Invoice INV-731. Supplier: Cedar Works. Subtotal: 80 EUR. Tax: 16 EUR. Total: 96 EUR. Payment due: not stated. Return those fields plus payment_due; use null when absent and preserve the identifier exactly.A labelled record with invoice_id, supplier, total, currency and payment_due, using null for the unstated date.Check that the total is taken from the Total line rather than the subtotal and that EUR is retained.