Gemma 3 4B IT is the instruction-tuned 4B-size edition of Google’s Gemma 3 vision-language family. IT identifies the version adapted to follow user instructions and conversation turns, rather than the pre-trained checkpoint used as a foundation for further training. The model receives text and images and generates text: an answer, summary, explanation, draft or code. Its visual input supports questions about supplied pictures and document pages; the output is not a newly generated image.
The model card assigns the 4B edition a 128K-token context shared by the request and generated answer. The remaining output allowance is the total context after subtracting input, not an additional independent 128K output budget. Its base image path normalizes images to 896×896 and encodes each view into 256 visual tokens. For a focused assistant, provide the relevant note or image, one clear question and a concise desired result. Keep the source text with the answer so an extracted name, date or label can be checked.
The family’s training material covers web text in over 140 languages, code, mathematics and images. The maker reports four trillion pre-training tokens for the 4B model. The technical report describes distillation during pre-training and instruction tuning, followed by reinforcement-learning-based post-training to improve helpfulness, reasoning, code, mathematics and instruction following. These are training characteristics of this size and family, not a guarantee that every answer is correct or that another Gemma checkpoint behaves identically.
The architecture is a decoder-only transformer with grouped-query attention, interleaving five local-attention layers for each global-attention layer. Local attention spans 1,024 tokens while global layers handle the wider context. A shared SigLIP vision encoder converts image information into inputs for the language model. The tokenizer is SentencePiece-based, with split digits, preserved whitespace and byte-level encoding. Use token counts from the actual tokenizer when planning a long input, and keep code indentation and identifiers intact when they carry meaning.
The report describes Pan and Scan for images whose shape or detail does not survive a simple square resize. It can create additional non-overlapping views that are resized for the vision encoder. That processing has its own input and computation implications. For a screenshot containing small labels, identify the region of interest and provide a readable view instead of relying on a tiny whole-screen picture. For several images, label their roles and ask for an answer linked to the correct source.
The maker’s native IT conversation format uses user and model turns, with explicit turn boundaries. It does not support a separate native system role; place role and formatting instructions in the initial user prompt. Use the matching formatter or processor rather than treating the IT checkpoint like plain text completion. For a support-note example, state the intended reader, facts to preserve and tone, then request a short reply plus any unresolved question. For visual extraction, define the fields and how an unreadable value should be represented.
Gemma’s intended uses include writing, summarization, conversational assistance, image-data extraction and educational exploration. A narrow 4B workflow can turn a service note into an action list, explain a visible diagram or answer a question about a provided passage. Ask for a result that is easy to inspect: an explanation tied to named elements or a small record with source-supported fields. Any tool lookup or code execution belongs to the surrounding application; generating a tool-shaped instruction is different from completing an external operation.
The maker notes limitations from training coverage, language nuance, factual errors and open-ended task complexity. Google also cautions that a plausible language pattern can produce an answer that misses an everyday real-world constraint. Images can also lose important detail through resizing, and a long context does not make an underspecified question clear. Review the result against the original material, validate requested structured fields and check calculations or proposed code independently. Evaluate the exact 4B IT model on the real language and document types of the task. Keep its size, IT identity and processing setup separate from the 12B, 27B and newer Gemma variants when recording results.
", "aboutText": "Gemma 3 4B IT is the instruction-tuned 4B-size edition of Google’s Gemma 3 vision-language family. IT identifies the version adapted to follow user instructions and conversation turns, rather than the pre-trained checkpoint used as a foundation for further training. The model receives text and images and generates text: an answer, summary, explanation, draft or code. Its visual input supports questions about supplied pictures and document pages; the output is not a newly generated image.\n\nThe model card assigns the 4B edition a 128K-token context shared by the request and generated answer. The remaining output allowance is the total context after subtracting input, not an additional independent 128K output budget. Its base image path normalizes images to 896×896 and encodes each view into 256 visual tokens. For a focused assistant, provide the relevant note or image, one clear question and a concise desired result. Keep the source text with the answer so an extracted name, date or label can be checked.\n\nThe family’s training material covers web text in over 140 languages, code, mathematics and images. The maker reports four trillion pre-training tokens for the 4B model. The technical report describes distillation during pre-training and instruction tuning, followed by reinforcement-learning-based post-training to improve helpfulness, reasoning, code, mathematics and instruction following. These are training characteristics of this size and family, not a guarantee that every answer is correct or that another Gemma checkpoint behaves identically.\n\nThe architecture is a decoder-only transformer with grouped-query attention, interleaving five local-attention layers for each global-attention layer. Local attention spans 1,024 tokens while global layers handle the wider context. A shared SigLIP vision encoder converts image information into inputs for the language model. The tokenizer is SentencePiece-based, with split digits, preserved whitespace and byte-level encoding. Use token counts from the actual tokenizer when planning a long input, and keep code indentation and identifiers intact when they carry meaning.\n\nThe report describes Pan and Scan for images whose shape or detail does not survive a simple square resize. It can create additional non-overlapping views that are resized for the vision encoder. That processing has its own input and computation implications. For a screenshot containing small labels, identify the region of interest and provide a readable view instead of relying on a tiny whole-screen picture. For several images, label their roles and ask for an answer linked to the correct source.\n\nThe maker’s native IT conversation format uses user and model turns, with explicit turn boundaries. It does not support a separate native system role; place role and formatting instructions in the initial user prompt. Use the matching formatter or processor rather than treating the IT checkpoint like plain text completion. For a support-note example, state the intended reader, facts to preserve and tone, then request a short reply plus any unresolved question. For visual extraction, define the fields and how an unreadable value should be represented.\n\nGemma’s intended uses include writing, summarization, conversational assistance, image-data extraction and educational exploration. A narrow 4B workflow can turn a service note into an action list, explain a visible diagram or answer a question about a provided passage. Ask for a result that is easy to inspect: an explanation tied to named elements or a small record with source-supported fields. Any tool lookup or code execution belongs to the surrounding application; generating a tool-shaped instruction is different from completing an external operation.\n\nThe maker notes limitations from training coverage, language nuance, factual errors and open-ended task complexity. Google also cautions that a plausible language pattern can produce an answer that misses an everyday real-world constraint. Images can also lose important detail through resizing, and a long context does not make an underspecified question clear. Review the result against the original material, validate requested structured fields and check calculations or proposed code independently. Evaluate the exact 4B IT model on the real language and document types of the task. Keep its size, IT identity and processing setup separate from the 12B, 27B and newer Gemma variants when recording results.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Explain a seed-growth note\n\nExplain this class note to a ten-year-old: 'Seeds need water and a suitable temperature to germinate. Light requirements differ between species.' Use three short sentences and one question the class could investigate. Preserve the species qualification rather than saying all seeds need light to germinate.A child-friendly three-sentence explanation followed by one investigation question.The explanation must retain the distinction between water, temperature and species-dependent light.google/gemma-3-4b-itGemma 3 4B IT is Google’s instruction-tuned 4B-size text-and-image model. It generates text for questions, summaries, writing and visual analysis, with a shared 128K context and family support for over 140 languages.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Gemma 3 4B IT is the instruction-tuned 4B-size edition of Google’s Gemma 3 vision-language family. IT identifies the version adapted to follow user instructions and conversation turns, rather than the pre-trained checkpoint used as a foundation for further training. The model receives text and images and generates text: an answer, summary, explanation, draft or code. Its visual input supports questions about supplied pictures and document pages; the output is not a newly generated image.
The model card assigns the 4B edition a 128K-token context shared by the request and generated answer. The remaining output allowance is the total context after subtracting input, not an additional independent 128K output budget. Its base image path normalizes images to 896×896 and encodes each view into 256 visual tokens. For a focused assistant, provide the relevant note or image, one clear question and a concise desired result. Keep the source text with the answer so an extracted name, date or label can be checked.
The family’s training material covers web text in over 140 languages, code, mathematics and images. The maker reports four trillion pre-training tokens for the 4B model. The technical report describes distillation during pre-training and instruction tuning, followed by reinforcement-learning-based post-training to improve helpfulness, reasoning, code, mathematics and instruction following. These are training characteristics of this size and family, not a guarantee that every answer is correct or that another Gemma checkpoint behaves identically.
The architecture is a decoder-only transformer with grouped-query attention, interleaving five local-attention layers for each global-attention layer. Local attention spans 1,024 tokens while global layers handle the wider context. A shared SigLIP vision encoder converts image information into inputs for the language model. The tokenizer is SentencePiece-based, with split digits, preserved whitespace and byte-level encoding. Use token counts from the actual tokenizer when planning a long input, and keep code indentation and identifiers intact when they carry meaning.
The report describes Pan and Scan for images whose shape or detail does not survive a simple square resize. It can create additional non-overlapping views that are resized for the vision encoder. That processing has its own input and computation implications. For a screenshot containing small labels, identify the region of interest and provide a readable view instead of relying on a tiny whole-screen picture. For several images, label their roles and ask for an answer linked to the correct source.
The maker’s native IT conversation format uses user and model turns, with explicit turn boundaries. It does not support a separate native system role; place role and formatting instructions in the initial user prompt. Use the matching formatter or processor rather than treating the IT checkpoint like plain text completion. For a support-note example, state the intended reader, facts to preserve and tone, then request a short reply plus any unresolved question. For visual extraction, define the fields and how an unreadable value should be represented.
Gemma’s intended uses include writing, summarization, conversational assistance, image-data extraction and educational exploration. A narrow 4B workflow can turn a service note into an action list, explain a visible diagram or answer a question about a provided passage. Ask for a result that is easy to inspect: an explanation tied to named elements or a small record with source-supported fields. Any tool lookup or code execution belongs to the surrounding application; generating a tool-shaped instruction is different from completing an external operation.
The maker notes limitations from training coverage, language nuance, factual errors and open-ended task complexity. Google also cautions that a plausible language pattern can produce an answer that misses an everyday real-world constraint. Images can also lose important detail through resizing, and a long context does not make an underspecified question clear. Review the result against the original material, validate requested structured fields and check calculations or proposed code independently. Evaluate the exact 4B IT model on the real language and document types of the task. Keep its size, IT identity and processing setup separate from the 12B, 27B and newer Gemma variants when recording results.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits google/gemma-3-4b-it | 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 $0.0001 / 1K tokens Output · Output tokens $0.0002 / 1K tokens Input · Cache read — / 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 $0.0001 / 1K tokens Output · Output tokens $0.0002 / 1K tokens Input · Cache read — / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model": "google/gemma-3-4b-it", "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='google/gemma-3-4b-it',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='google/gemma-3-4b-it',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 google/gemma-3-4b-it
Use the method paths shown above.
google/gemma-3-4b-itGemma 3 4B IT is Google’s instruction-tuned 4B-size text-and-image model. It generates text for questions, summaries, writing and visual analysis, with a shared 128K context and family support for over 140 languages.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Gemma 3 4B IT is the instruction-tuned 4B-size edition of Google’s Gemma 3 vision-language family. IT identifies the version adapted to follow user instructions and conversation turns, rather than the pre-trained checkpoint used as a foundation for further training. The model receives text and images and generates text: an answer, summary, explanation, draft or code. Its visual input supports questions about supplied pictures and document pages; the output is not a newly generated image.
The model card assigns the 4B edition a 128K-token context shared by the request and generated answer. The remaining output allowance is the total context after subtracting input, not an additional independent 128K output budget. Its base image path normalizes images to 896×896 and encodes each view into 256 visual tokens. For a focused assistant, provide the relevant note or image, one clear question and a concise desired result. Keep the source text with the answer so an extracted name, date or label can be checked.
The family’s training material covers web text in over 140 languages, code, mathematics and images. The maker reports four trillion pre-training tokens for the 4B model. The technical report describes distillation during pre-training and instruction tuning, followed by reinforcement-learning-based post-training to improve helpfulness, reasoning, code, mathematics and instruction following. These are training characteristics of this size and family, not a guarantee that every answer is correct or that another Gemma checkpoint behaves identically.
The architecture is a decoder-only transformer with grouped-query attention, interleaving five local-attention layers for each global-attention layer. Local attention spans 1,024 tokens while global layers handle the wider context. A shared SigLIP vision encoder converts image information into inputs for the language model. The tokenizer is SentencePiece-based, with split digits, preserved whitespace and byte-level encoding. Use token counts from the actual tokenizer when planning a long input, and keep code indentation and identifiers intact when they carry meaning.
The report describes Pan and Scan for images whose shape or detail does not survive a simple square resize. It can create additional non-overlapping views that are resized for the vision encoder. That processing has its own input and computation implications. For a screenshot containing small labels, identify the region of interest and provide a readable view instead of relying on a tiny whole-screen picture. For several images, label their roles and ask for an answer linked to the correct source.
The maker’s native IT conversation format uses user and model turns, with explicit turn boundaries. It does not support a separate native system role; place role and formatting instructions in the initial user prompt. Use the matching formatter or processor rather than treating the IT checkpoint like plain text completion. For a support-note example, state the intended reader, facts to preserve and tone, then request a short reply plus any unresolved question. For visual extraction, define the fields and how an unreadable value should be represented.
Gemma’s intended uses include writing, summarization, conversational assistance, image-data extraction and educational exploration. A narrow 4B workflow can turn a service note into an action list, explain a visible diagram or answer a question about a provided passage. Ask for a result that is easy to inspect: an explanation tied to named elements or a small record with source-supported fields. Any tool lookup or code execution belongs to the surrounding application; generating a tool-shaped instruction is different from completing an external operation.
The maker notes limitations from training coverage, language nuance, factual errors and open-ended task complexity. Google also cautions that a plausible language pattern can produce an answer that misses an everyday real-world constraint. Images can also lose important detail through resizing, and a long context does not make an underspecified question clear. Review the result against the original material, validate requested structured fields and check calculations or proposed code independently. Evaluate the exact 4B IT model on the real language and document types of the task. Keep its size, IT identity and processing setup separate from the 12B, 27B and newer Gemma variants when recording results.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits google/gemma-3-4b-it | 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 $0.0001 / 1K tokens Output · Output tokens $0.0002 / 1K tokens Input · Cache read — / 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 $0.0001 / 1K tokens Output · Output tokens $0.0002 / 1K tokens Input · Cache read — / 1K tokens Input · Cache write — / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "google/gemma-3-4b-it",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${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': 'google/gemma-3-4b-it', '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': 'google/gemma-3-4b-it', '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 google/gemma-3-4b-it
Use the method paths shown above.
Technical example in English
Explain a seed-growth note\n\nExplain this class note to a ten-year-old: 'Seeds need water and a suitable temperature to germinate. Light requirements differ between species.' Use three short sentences and one question the class could investigate. Preserve the species qualification rather than saying all seeds need light to germinate.A child-friendly three-sentence explanation followed by one investigation question.The explanation must retain the distinction between water, temperature and species-dependent light.