Mistral Small 3.1 24B Instruct 2503 builds on Small 3 2501 by adding visual understanding and extending long-context support to 128k tokens. It has 24 billion parameters and is instruction-tuned from the separately named Small 3.1 Base release. The model accepts text and images and produces text, including an explanation, code or a structured record.
Key capabilities
Vision lets the model analyze visible material alongside the user's question. A document image can supply the field or statement to inspect, a diagram can supply the relationship to explain, and a chart can supply the data behind a comparison. Keep labels, units and relevant surrounding content readable so the question refers to evidence actually present in the image.
The maker lists multilingual support across dozens of languages, including Russian, English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Serbian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese and Farsi. Define the final language, reader and terminology for a multilingual answer. Names, quoted labels, variable identifiers and units can remain unchanged while the surrounding prose is adapted.
The card also describes conversational and reasoning capabilities, native function calling, JSON output and adherence to system prompts. Its tokenizer is Tekken, with a vocabulary labeled 131k. These are features of the named instruction-tuned release; its Base edition serves a different role and later Small 3.2 and Small 4 releases retain their own changes and modes.
Document, coding and mathematical preparation
For a long-document task, identify the documents and ask one defined question. A comparison can request a requirement matrix with the statement, its source section and any disagreement between documents. A summary can request only the material relevant to a named change. The long context keeps those sources together, while the final answer can remain concise.
For a coding explanation, provide the function and the input whose behavior should be traced. Include relevant helpers or declarations and ask how the state changes before the value is returned. For a proposed modification, state the requirement, the public interface to preserve and the expected output. Request the revised code separately from its explanation and examples.
For a mathematical task, define the symbols and domain assumptions before asking for a derivation. Specify whether the final result should be exact, rounded or a proof of a claim. If the condition appears in an image, keep the formula and associated labels clear and ask the model to preserve that notation in its answer. These preparations turn the model's documented programming and math reasoning focus into an inspectable artifact.
Instruction format and recommended usage
The maker publishes a V7-Tekken conversation template that distinguishes system, user and assistant roles and recommends using the matching tokenizer implementation as the source of truth. A system instruction should define the assistant's role and response rules, while the user message contains the actual task and source material. Correct conversation formatting is different from placing role-like labels inside a quoted document.
The card recommends a relatively low temperature, such as 0.15, and a system prompt tailored to the use case. Define the output fields or sections and the conditions that must survive rewriting. For repeated extraction, ask for JSON with specified names and types and a representation for values missing from the source. An explanatory paragraph can be returned separately when a person needs to inspect the interpretation.
Function-calling workflow
The model can select a function from tools described by the application and generate its arguments. The application executes the function, returns the result with the corresponding tool call, and the model uses that result in a subsequent answer. The maker's example shows this conversation cycle, including prior tool use and a new user question. A proposed lookup is therefore distinct from the data actually returned by the lookup.
A practical assistant might first extract a question from a document and then use a supplied reference function to obtain a missing condition. Describe the function clearly, include its required arguments and keep the returned result associated with the originating operation. JSON organizes a text result; function calling requests a connected operation rather than silently executing code or reading a database.
Substantive limits
The documented visual capability reads images; it does not generate them. The released card's assistant guidance also distinguishes this model from audio and video transcription, so a spoken or video question needs its relevant content supplied through an appropriate separate preparation step. Web access and external actions likewise come from connected tools, not the text-and-image model alone.
The 128k context concerns the task's input and generated conversation, while the card's 512-token generation example is only an example setting. No separate hard answer-length capacity is specified in the sources read. Preserve the exact Instruct 2503 edition, its conversation format and the requested explanation, record or code artifact.
", "aboutText": "Mistral Small 3.1 24B Instruct 2503 builds on Small 3 2501 by adding visual understanding and extending long-context support to 128k tokens. It has 24 billion parameters and is instruction-tuned from the separately named Small 3.1 Base release. The model accepts text and images and produces text, including an explanation, code or a structured record.\n\nKey capabilities\n\nVision lets the model analyze visible material alongside the user's question. A document image can supply the field or statement to inspect, a diagram can supply the relationship to explain, and a chart can supply the data behind a comparison. Keep labels, units and relevant surrounding content readable so the question refers to evidence actually present in the image.\n\nThe maker lists multilingual support across dozens of languages, including Russian, English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Serbian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese and Farsi. Define the final language, reader and terminology for a multilingual answer. Names, quoted labels, variable identifiers and units can remain unchanged while the surrounding prose is adapted.\n\nThe card also describes conversational and reasoning capabilities, native function calling, JSON output and adherence to system prompts. Its tokenizer is Tekken, with a vocabulary labeled 131k. These are features of the named instruction-tuned release; its Base edition serves a different role and later Small 3.2 and Small 4 releases retain their own changes and modes.\n\nDocument, coding and mathematical preparation\n\nFor a long-document task, identify the documents and ask one defined question. A comparison can request a requirement matrix with the statement, its source section and any disagreement between documents. A summary can request only the material relevant to a named change. The long context keeps those sources together, while the final answer can remain concise.\n\nFor a coding explanation, provide the function and the input whose behavior should be traced. Include relevant helpers or declarations and ask how the state changes before the value is returned. For a proposed modification, state the requirement, the public interface to preserve and the expected output. Request the revised code separately from its explanation and examples.\n\nFor a mathematical task, define the symbols and domain assumptions before asking for a derivation. Specify whether the final result should be exact, rounded or a proof of a claim. If the condition appears in an image, keep the formula and associated labels clear and ask the model to preserve that notation in its answer. These preparations turn the model's documented programming and math reasoning focus into an inspectable artifact.\n\nInstruction format and recommended usage\n\nThe maker publishes a V7-Tekken conversation template that distinguishes system, user and assistant roles and recommends using the matching tokenizer implementation as the source of truth. A system instruction should define the assistant's role and response rules, while the user message contains the actual task and source material. Correct conversation formatting is different from placing role-like labels inside a quoted document.\n\nThe card recommends a relatively low temperature, such as 0.15, and a system prompt tailored to the use case. Define the output fields or sections and the conditions that must survive rewriting. For repeated extraction, ask for JSON with specified names and types and a representation for values missing from the source. An explanatory paragraph can be returned separately when a person needs to inspect the interpretation.\n\nFunction-calling workflow\n\nThe model can select a function from tools described by the application and generate its arguments. The application executes the function, returns the result with the corresponding tool call, and the model uses that result in a subsequent answer. The maker's example shows this conversation cycle, including prior tool use and a new user question. A proposed lookup is therefore distinct from the data actually returned by the lookup.\n\nA practical assistant might first extract a question from a document and then use a supplied reference function to obtain a missing condition. Describe the function clearly, include its required arguments and keep the returned result associated with the originating operation. JSON organizes a text result; function calling requests a connected operation rather than silently executing code or reading a database.\n\nSubstantive limits\n\nThe documented visual capability reads images; it does not generate them. The released card's assistant guidance also distinguishes this model from audio and video transcription, so a spoken or video question needs its relevant content supplied through an appropriate separate preparation step. Web access and external actions likewise come from connected tools, not the text-and-image model alone.\n\nThe 128k context concerns the task's input and generated conversation, while the card's 512-token generation example is only an example setting. No separate hard answer-length capacity is specified in the sources read. Preserve the exact Instruct 2503 edition, its conversation format and the requested explanation, record or code artifact.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Trace an accumulator function\n\nTrace this Python function on [3, -1, 4]: 'def running_sum(xs):\n total = 0\n out = []\n for x in xs:\n total += x\n out.append(total)\n return out'. Give a row for each iteration showing x, total and out, then the return value. Explain in one sentence why out differs from doubling each element.A three-row state trace, the final list and one sentence distinguishing cumulative addition from elementwise multiplication.Each row uses the prior total, including the negative input, rather than restarting the accumulator.mistralai/mistral-small-3.1-24b-instructMistral Small 3.1 24B Instruct 2503 adds image understanding and a 128k maker context to the Small 3 line. It combines multilingual text, programming and mathematical reasoning with system instructions, JSON and native function calling in a distinct instruction-tuned release.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Mistral Small 3.1 24B Instruct 2503 builds on Small 3 2501 by adding visual understanding and extending long-context support to 128k tokens. It has 24 billion parameters and is instruction-tuned from the separately named Small 3.1 Base release. The model accepts text and images and produces text, including an explanation, code or a structured record.
Key capabilities
Vision lets the model analyze visible material alongside the user's question. A document image can supply the field or statement to inspect, a diagram can supply the relationship to explain, and a chart can supply the data behind a comparison. Keep labels, units and relevant surrounding content readable so the question refers to evidence actually present in the image.
The maker lists multilingual support across dozens of languages, including Russian, English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Serbian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese and Farsi. Define the final language, reader and terminology for a multilingual answer. Names, quoted labels, variable identifiers and units can remain unchanged while the surrounding prose is adapted.
The card also describes conversational and reasoning capabilities, native function calling, JSON output and adherence to system prompts. Its tokenizer is Tekken, with a vocabulary labeled 131k. These are features of the named instruction-tuned release; its Base edition serves a different role and later Small 3.2 and Small 4 releases retain their own changes and modes.
Document, coding and mathematical preparation
For a long-document task, identify the documents and ask one defined question. A comparison can request a requirement matrix with the statement, its source section and any disagreement between documents. A summary can request only the material relevant to a named change. The long context keeps those sources together, while the final answer can remain concise.
For a coding explanation, provide the function and the input whose behavior should be traced. Include relevant helpers or declarations and ask how the state changes before the value is returned. For a proposed modification, state the requirement, the public interface to preserve and the expected output. Request the revised code separately from its explanation and examples.
For a mathematical task, define the symbols and domain assumptions before asking for a derivation. Specify whether the final result should be exact, rounded or a proof of a claim. If the condition appears in an image, keep the formula and associated labels clear and ask the model to preserve that notation in its answer. These preparations turn the model's documented programming and math reasoning focus into an inspectable artifact.
Instruction format and recommended usage
The maker publishes a V7-Tekken conversation template that distinguishes system, user and assistant roles and recommends using the matching tokenizer implementation as the source of truth. A system instruction should define the assistant's role and response rules, while the user message contains the actual task and source material. Correct conversation formatting is different from placing role-like labels inside a quoted document.
The card recommends a relatively low temperature, such as 0.15, and a system prompt tailored to the use case. Define the output fields or sections and the conditions that must survive rewriting. For repeated extraction, ask for JSON with specified names and types and a representation for values missing from the source. An explanatory paragraph can be returned separately when a person needs to inspect the interpretation.
Function-calling workflow
The model can select a function from tools described by the application and generate its arguments. The application executes the function, returns the result with the corresponding tool call, and the model uses that result in a subsequent answer. The maker's example shows this conversation cycle, including prior tool use and a new user question. A proposed lookup is therefore distinct from the data actually returned by the lookup.
A practical assistant might first extract a question from a document and then use a supplied reference function to obtain a missing condition. Describe the function clearly, include its required arguments and keep the returned result associated with the originating operation. JSON organizes a text result; function calling requests a connected operation rather than silently executing code or reading a database.
Substantive limits
The documented visual capability reads images; it does not generate them. The released card's assistant guidance also distinguishes this model from audio and video transcription, so a spoken or video question needs its relevant content supplied through an appropriate separate preparation step. Web access and external actions likewise come from connected tools, not the text-and-image model alone.
The 128k context concerns the task's input and generated conversation, while the card's 512-token generation example is only an example setting. No separate hard answer-length capacity is specified in the sources read. Preserve the exact Instruct 2503 edition, its conversation format and the requested explanation, record or code artifact.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits mistralai/mistral-small-3.1-24b-instruct | 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 $0.0011 / 1K tokens Output · Output tokens $0.00111 / 1K tokens Input · Cache read $0 / 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 $0.0011 / 1K tokens Output · Output tokens $0.00111 / 1K tokens Input · Cache read $0 / 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/mistral-small-3.1-24b-instruct", "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/mistral-small-3.1-24b-instruct',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/mistral-small-3.1-24b-instruct',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/mistral-small-3.1-24b-instruct
Use the method paths shown above.
mistralai/mistral-small-3.1-24b-instructMistral Small 3.1 24B Instruct 2503 adds image understanding and a 128k maker context to the Small 3 line. It combines multilingual text, programming and mathematical reasoning with system instructions, JSON and native function calling in a distinct instruction-tuned release.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Mistral Small 3.1 24B Instruct 2503 builds on Small 3 2501 by adding visual understanding and extending long-context support to 128k tokens. It has 24 billion parameters and is instruction-tuned from the separately named Small 3.1 Base release. The model accepts text and images and produces text, including an explanation, code or a structured record.
Key capabilities
Vision lets the model analyze visible material alongside the user's question. A document image can supply the field or statement to inspect, a diagram can supply the relationship to explain, and a chart can supply the data behind a comparison. Keep labels, units and relevant surrounding content readable so the question refers to evidence actually present in the image.
The maker lists multilingual support across dozens of languages, including Russian, English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Serbian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese and Farsi. Define the final language, reader and terminology for a multilingual answer. Names, quoted labels, variable identifiers and units can remain unchanged while the surrounding prose is adapted.
The card also describes conversational and reasoning capabilities, native function calling, JSON output and adherence to system prompts. Its tokenizer is Tekken, with a vocabulary labeled 131k. These are features of the named instruction-tuned release; its Base edition serves a different role and later Small 3.2 and Small 4 releases retain their own changes and modes.
Document, coding and mathematical preparation
For a long-document task, identify the documents and ask one defined question. A comparison can request a requirement matrix with the statement, its source section and any disagreement between documents. A summary can request only the material relevant to a named change. The long context keeps those sources together, while the final answer can remain concise.
For a coding explanation, provide the function and the input whose behavior should be traced. Include relevant helpers or declarations and ask how the state changes before the value is returned. For a proposed modification, state the requirement, the public interface to preserve and the expected output. Request the revised code separately from its explanation and examples.
For a mathematical task, define the symbols and domain assumptions before asking for a derivation. Specify whether the final result should be exact, rounded or a proof of a claim. If the condition appears in an image, keep the formula and associated labels clear and ask the model to preserve that notation in its answer. These preparations turn the model's documented programming and math reasoning focus into an inspectable artifact.
Instruction format and recommended usage
The maker publishes a V7-Tekken conversation template that distinguishes system, user and assistant roles and recommends using the matching tokenizer implementation as the source of truth. A system instruction should define the assistant's role and response rules, while the user message contains the actual task and source material. Correct conversation formatting is different from placing role-like labels inside a quoted document.
The card recommends a relatively low temperature, such as 0.15, and a system prompt tailored to the use case. Define the output fields or sections and the conditions that must survive rewriting. For repeated extraction, ask for JSON with specified names and types and a representation for values missing from the source. An explanatory paragraph can be returned separately when a person needs to inspect the interpretation.
Function-calling workflow
The model can select a function from tools described by the application and generate its arguments. The application executes the function, returns the result with the corresponding tool call, and the model uses that result in a subsequent answer. The maker's example shows this conversation cycle, including prior tool use and a new user question. A proposed lookup is therefore distinct from the data actually returned by the lookup.
A practical assistant might first extract a question from a document and then use a supplied reference function to obtain a missing condition. Describe the function clearly, include its required arguments and keep the returned result associated with the originating operation. JSON organizes a text result; function calling requests a connected operation rather than silently executing code or reading a database.
Substantive limits
The documented visual capability reads images; it does not generate them. The released card's assistant guidance also distinguishes this model from audio and video transcription, so a spoken or video question needs its relevant content supplied through an appropriate separate preparation step. Web access and external actions likewise come from connected tools, not the text-and-image model alone.
The 128k context concerns the task's input and generated conversation, while the card's 512-token generation example is only an example setting. No separate hard answer-length capacity is specified in the sources read. Preserve the exact Instruct 2503 edition, its conversation format and the requested explanation, record or code artifact.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits mistralai/mistral-small-3.1-24b-instruct | 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 $0.0011 / 1K tokens Output · Output tokens $0.00111 / 1K tokens Input · Cache read $0 / 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 $0.0011 / 1K tokens Output · Output tokens $0.00111 / 1K tokens Input · Cache read $0 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "mistralai/mistral-small-3.1-24b-instruct",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${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/mistral-small-3.1-24b-instruct', '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/mistral-small-3.1-24b-instruct', '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/mistral-small-3.1-24b-instruct
Use the method paths shown above.
Technical example in English
Trace an accumulator function\n\nTrace this Python function on [3, -1, 4]: 'def running_sum(xs):\n total = 0\n out = []\n for x in xs:\n total += x\n out.append(total)\n return out'. Give a row for each iteration showing x, total and out, then the return value. Explain in one sentence why out differs from doubling each element.A three-row state trace, the final list and one sentence distinguishing cumulative addition from elementwise multiplication.Each row uses the prior total, including the negative input, rather than restarting the accumulator.