Qwen3.5 highlights: the 397B-A17B checkpoint
Qwen3.5-397B-A17B is the post-trained vision-language mixture-of-experts model in the Qwen3.5 family. It accepts text, images and video and generates text: explanations, code proposals, analysis and requested document artifacts. The family uses early text–vision fusion and reinforcement learning across progressively complex tasks to develop reasoning, coding, agents and visual understanding. Its documented linguistic coverage is 201 languages and dialects, rather than a promise of identical reliability for every language or specialist field.
The maker separately describes managed Qwen3.5-Plus as a corresponding offering with additional production features. Preserve this open checkpoint’s own identity, context and generation recipe instead of importing every managed feature or later Qwen control. The exact 397B-A17B edition remains distinct from smaller dense and expert models in the same family.
Model overview
The model has 397 billion parameters in total and activates 17 billion. Its 60-layer hybrid layout repeats three Gated DeltaNet expert layers followed by one gated-attention expert layer. The maker lists 512 routed experts, selecting ten routed experts together with one shared expert. A vision encoder and multi-token prediction are part of the model. Total parameters, activated parameters and the complete checkpoint’s storage size are different concepts; these counts do not establish a fixed response time.
The native context is 262,144 tokens, with conditional extension up to 1,010,000. Text, represented images or video, conversation history, reasoning and the final answer share the configured sequence allowance. Use the actual configured budget for a task rather than assume that the largest document and largest answer fit together automatically.
Preparing text and visual evidence
For a technical project, supply the specification, relevant code and diagrams and define the artifact to produce. Ask for an implementation proposal organized by affected components and supported by the supplied requirements. For a document comparison, preserve section identifiers, table labels and units and request a findings table with evidence for each row. For a multilingual report, specify the target language and terms that should stay unchanged.
For an image question, preserve small labels and provide definitions not present in the picture. For a video question, identify the relevant event and preserve frame order and timing. Compare the returned visual details, quotations and event descriptions with the original images, passages and timestamps. The result remains textual analysis or code; the model’s visual input does not imply image, video or speech generation.
Thinking and conversation history
Thinking is enabled by default and precedes the final response. The maker’s enable_thinking setting can disable it for a direct answer. The older /think and /nothink soft prompt switches are not officially supported for Qwen3.5. Keep the final artifact easy to locate after the reasoning, such as a delimited conclusion, evidence table or separate code proposal.
The card recommends retaining final answers without thinking content in historical assistant messages. That differs from historical reasoning-preservation options and defaults introduced in later releases. Follow this exact checkpoint’s template and keep corrections or tool observations associated with the task they update.
Agentic usage
Qwen documents tool calling and recommends Qwen-Agent with configured or MCP-connected tools, and Qwen Code for coding-agent workflows. State the permitted operations, available evidence and observable completion criteria. An application executes tools and supplies their results; the model can use those observations to revise its plan or explanation. A proposed command or patch remains distinct from an action actually completed in the environment.
Ultra-long text and video
The maker’s extension uses YaRN with the checkpoint’s own interleaved positional settings. A factor of four targets the roughly one-million-token workflow; a factor of two is suggested for 524,288-token workloads. Static scaling can affect shorter-text performance, so it should follow a real need for larger context. Organize a long input by document, page or code identifier and leave room for the result.
The supported video example uses two frames per second by default. For hour-scale material, the maker recommends a larger processor budget corresponding to roughly 224K video tokens for higher-rate sampling. That is a configuration-dependent recipe, not a universal media-duration, complete-recall or timestamp-precision guarantee.
Generation guidance and substantive limits
This checkpoint’s Best Practices recommend temperature 0.6, top_p 0.95, top_k 20 and min_p 0 for thinking, and temperature 0.7, top_p 0.8, top_k 20 and min_p 0 for direct responses. These exact recommendations differ from some smaller Qwen3.5 editions. Where supported, a presence penalty between zero and two can reduce repetition, but higher values may mix languages or slightly reduce quality.
Recommended generation allowances are 32,768 tokens for most queries and 81,920 for unusually complex mathematical or programming work. They are recommendations, not independent hard output ceilings. Specify the required answer format and check calculations, extracted details and proposed code against the source. The useful output is a concrete textual artifact whose evidence and assumptions can be reviewed.
", "aboutText": "Qwen3.5 highlights: the 397B-A17B checkpoint\n\nQwen3.5-397B-A17B is the post-trained vision-language mixture-of-experts model in the Qwen3.5 family. It accepts text, images and video and generates text: explanations, code proposals, analysis and requested document artifacts. The family uses early text–vision fusion and reinforcement learning across progressively complex tasks to develop reasoning, coding, agents and visual understanding. Its documented linguistic coverage is 201 languages and dialects, rather than a promise of identical reliability for every language or specialist field.\n\nThe maker separately describes managed Qwen3.5-Plus as a corresponding offering with additional production features. Preserve this open checkpoint’s own identity, context and generation recipe instead of importing every managed feature or later Qwen control. The exact 397B-A17B edition remains distinct from smaller dense and expert models in the same family.\n\nModel overview\n\nThe model has 397 billion parameters in total and activates 17 billion. Its 60-layer hybrid layout repeats three Gated DeltaNet expert layers followed by one gated-attention expert layer. The maker lists 512 routed experts, selecting ten routed experts together with one shared expert. A vision encoder and multi-token prediction are part of the model. Total parameters, activated parameters and the complete checkpoint’s storage size are different concepts; these counts do not establish a fixed response time.\n\nThe native context is 262,144 tokens, with conditional extension up to 1,010,000. Text, represented images or video, conversation history, reasoning and the final answer share the configured sequence allowance. Use the actual configured budget for a task rather than assume that the largest document and largest answer fit together automatically.\n\nPreparing text and visual evidence\n\nFor a technical project, supply the specification, relevant code and diagrams and define the artifact to produce. Ask for an implementation proposal organized by affected components and supported by the supplied requirements. For a document comparison, preserve section identifiers, table labels and units and request a findings table with evidence for each row. For a multilingual report, specify the target language and terms that should stay unchanged.\n\nFor an image question, preserve small labels and provide definitions not present in the picture. For a video question, identify the relevant event and preserve frame order and timing. Compare the returned visual details, quotations and event descriptions with the original images, passages and timestamps. The result remains textual analysis or code; the model’s visual input does not imply image, video or speech generation.\n\nThinking and conversation history\n\nThinking is enabled by default and precedes the final response. The maker’s enable_thinking setting can disable it for a direct answer. The older /think and /nothink soft prompt switches are not officially supported for Qwen3.5. Keep the final artifact easy to locate after the reasoning, such as a delimited conclusion, evidence table or separate code proposal.\n\nThe card recommends retaining final answers without thinking content in historical assistant messages. That differs from historical reasoning-preservation options and defaults introduced in later releases. Follow this exact checkpoint’s template and keep corrections or tool observations associated with the task they update.\n\nAgentic usage\n\nQwen documents tool calling and recommends Qwen-Agent with configured or MCP-connected tools, and Qwen Code for coding-agent workflows. State the permitted operations, available evidence and observable completion criteria. An application executes tools and supplies their results; the model can use those observations to revise its plan or explanation. A proposed command or patch remains distinct from an action actually completed in the environment.\n\nUltra-long text and video\n\nThe maker’s extension uses YaRN with the checkpoint’s own interleaved positional settings. A factor of four targets the roughly one-million-token workflow; a factor of two is suggested for 524,288-token workloads. Static scaling can affect shorter-text performance, so it should follow a real need for larger context. Organize a long input by document, page or code identifier and leave room for the result.\n\nThe supported video example uses two frames per second by default. For hour-scale material, the maker recommends a larger processor budget corresponding to roughly 224K video tokens for higher-rate sampling. That is a configuration-dependent recipe, not a universal media-duration, complete-recall or timestamp-precision guarantee.\n\nGeneration guidance and substantive limits\n\nThis checkpoint’s Best Practices recommend temperature 0.6, top_p 0.95, top_k 20 and min_p 0 for thinking, and temperature 0.7, top_p 0.8, top_k 20 and min_p 0 for direct responses. These exact recommendations differ from some smaller Qwen3.5 editions. Where supported, a presence penalty between zero and two can reduce repetition, but higher values may mix languages or slightly reduce quality.\n\nRecommended generation allowances are 32,768 tokens for most queries and 81,920 for unusually complex mathematical or programming work. They are recommendations, not independent hard output ceilings. Specify the required answer format and check calculations, extracted details and proposed code against the source. The useful output is a concrete textual artifact whose evidence and assumptions can be reviewed.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Reconcile a specification and implementation plan\n\nCompare spec S1: users may delete only their own drafts, and published records cannot be deleted. Plan P1: show Delete for every record owned by the user, regardless of status. Return the exact mismatch, a corrected eligibility rule and test cases for an owned draft, owned published record and another user’s draft.An evidence-linked mismatch, a corrected rule and three acceptance-test descriptions.Check that both ownership and draft status are required rather than treating either condition as sufficient.qwen/qwen3.5-397b-a17bQwen3.5-397B-A17B is Qwen’s vision-language mixture-of-experts model with 397 billion total and 17 billion activated parameters. It accepts text, images and video and returns text for reasoning, coding and agent tasks, with selectable thinking and a native 262,144-token context.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Qwen3.5 highlights: the 397B-A17B checkpoint
Qwen3.5-397B-A17B is the post-trained vision-language mixture-of-experts model in the Qwen3.5 family. It accepts text, images and video and generates text: explanations, code proposals, analysis and requested document artifacts. The family uses early text–vision fusion and reinforcement learning across progressively complex tasks to develop reasoning, coding, agents and visual understanding. Its documented linguistic coverage is 201 languages and dialects, rather than a promise of identical reliability for every language or specialist field.
The maker separately describes managed Qwen3.5-Plus as a corresponding offering with additional production features. Preserve this open checkpoint’s own identity, context and generation recipe instead of importing every managed feature or later Qwen control. The exact 397B-A17B edition remains distinct from smaller dense and expert models in the same family.
Model overview
The model has 397 billion parameters in total and activates 17 billion. Its 60-layer hybrid layout repeats three Gated DeltaNet expert layers followed by one gated-attention expert layer. The maker lists 512 routed experts, selecting ten routed experts together with one shared expert. A vision encoder and multi-token prediction are part of the model. Total parameters, activated parameters and the complete checkpoint’s storage size are different concepts; these counts do not establish a fixed response time.
The native context is 262,144 tokens, with conditional extension up to 1,010,000. Text, represented images or video, conversation history, reasoning and the final answer share the configured sequence allowance. Use the actual configured budget for a task rather than assume that the largest document and largest answer fit together automatically.
Preparing text and visual evidence
For a technical project, supply the specification, relevant code and diagrams and define the artifact to produce. Ask for an implementation proposal organized by affected components and supported by the supplied requirements. For a document comparison, preserve section identifiers, table labels and units and request a findings table with evidence for each row. For a multilingual report, specify the target language and terms that should stay unchanged.
For an image question, preserve small labels and provide definitions not present in the picture. For a video question, identify the relevant event and preserve frame order and timing. Compare the returned visual details, quotations and event descriptions with the original images, passages and timestamps. The result remains textual analysis or code; the model’s visual input does not imply image, video or speech generation.
Thinking and conversation history
Thinking is enabled by default and precedes the final response. The maker’s enable_thinking setting can disable it for a direct answer. The older /think and /nothink soft prompt switches are not officially supported for Qwen3.5. Keep the final artifact easy to locate after the reasoning, such as a delimited conclusion, evidence table or separate code proposal.
The card recommends retaining final answers without thinking content in historical assistant messages. That differs from historical reasoning-preservation options and defaults introduced in later releases. Follow this exact checkpoint’s template and keep corrections or tool observations associated with the task they update.
Agentic usage
Qwen documents tool calling and recommends Qwen-Agent with configured or MCP-connected tools, and Qwen Code for coding-agent workflows. State the permitted operations, available evidence and observable completion criteria. An application executes tools and supplies their results; the model can use those observations to revise its plan or explanation. A proposed command or patch remains distinct from an action actually completed in the environment.
Ultra-long text and video
The maker’s extension uses YaRN with the checkpoint’s own interleaved positional settings. A factor of four targets the roughly one-million-token workflow; a factor of two is suggested for 524,288-token workloads. Static scaling can affect shorter-text performance, so it should follow a real need for larger context. Organize a long input by document, page or code identifier and leave room for the result.
The supported video example uses two frames per second by default. For hour-scale material, the maker recommends a larger processor budget corresponding to roughly 224K video tokens for higher-rate sampling. That is a configuration-dependent recipe, not a universal media-duration, complete-recall or timestamp-precision guarantee.
Generation guidance and substantive limits
This checkpoint’s Best Practices recommend temperature 0.6, top_p 0.95, top_k 20 and min_p 0 for thinking, and temperature 0.7, top_p 0.8, top_k 20 and min_p 0 for direct responses. These exact recommendations differ from some smaller Qwen3.5 editions. Where supported, a presence penalty between zero and two can reduce repetition, but higher values may mix languages or slightly reduce quality.
Recommended generation allowances are 32,768 tokens for most queries and 81,920 for unusually complex mathematical or programming work. They are recommendations, not independent hard output ceilings. Specify the required answer format and check calculations, extracted details and proposed code against the source. The useful output is a concrete textual artifact whose evidence and assumptions can be reviewed.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits qwen/qwen3.5-397b-a17b | 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 $0.00078 / 1K tokens Output · Output tokens $0.00468 / 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 $0.00078 / 1K tokens Output · Output tokens $0.00468 / 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": "qwen/qwen3.5-397b-a17b", "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='qwen/qwen3.5-397b-a17b',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='qwen/qwen3.5-397b-a17b',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 qwen/qwen3.5-397b-a17b
Use the method paths shown above.
qwen/qwen3.5-397b-a17bQwen3.5-397B-A17B is Qwen’s vision-language mixture-of-experts model with 397 billion total and 17 billion activated parameters. It accepts text, images and video and returns text for reasoning, coding and agent tasks, with selectable thinking and a native 262,144-token context.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Qwen3.5 highlights: the 397B-A17B checkpoint
Qwen3.5-397B-A17B is the post-trained vision-language mixture-of-experts model in the Qwen3.5 family. It accepts text, images and video and generates text: explanations, code proposals, analysis and requested document artifacts. The family uses early text–vision fusion and reinforcement learning across progressively complex tasks to develop reasoning, coding, agents and visual understanding. Its documented linguistic coverage is 201 languages and dialects, rather than a promise of identical reliability for every language or specialist field.
The maker separately describes managed Qwen3.5-Plus as a corresponding offering with additional production features. Preserve this open checkpoint’s own identity, context and generation recipe instead of importing every managed feature or later Qwen control. The exact 397B-A17B edition remains distinct from smaller dense and expert models in the same family.
Model overview
The model has 397 billion parameters in total and activates 17 billion. Its 60-layer hybrid layout repeats three Gated DeltaNet expert layers followed by one gated-attention expert layer. The maker lists 512 routed experts, selecting ten routed experts together with one shared expert. A vision encoder and multi-token prediction are part of the model. Total parameters, activated parameters and the complete checkpoint’s storage size are different concepts; these counts do not establish a fixed response time.
The native context is 262,144 tokens, with conditional extension up to 1,010,000. Text, represented images or video, conversation history, reasoning and the final answer share the configured sequence allowance. Use the actual configured budget for a task rather than assume that the largest document and largest answer fit together automatically.
Preparing text and visual evidence
For a technical project, supply the specification, relevant code and diagrams and define the artifact to produce. Ask for an implementation proposal organized by affected components and supported by the supplied requirements. For a document comparison, preserve section identifiers, table labels and units and request a findings table with evidence for each row. For a multilingual report, specify the target language and terms that should stay unchanged.
For an image question, preserve small labels and provide definitions not present in the picture. For a video question, identify the relevant event and preserve frame order and timing. Compare the returned visual details, quotations and event descriptions with the original images, passages and timestamps. The result remains textual analysis or code; the model’s visual input does not imply image, video or speech generation.
Thinking and conversation history
Thinking is enabled by default and precedes the final response. The maker’s enable_thinking setting can disable it for a direct answer. The older /think and /nothink soft prompt switches are not officially supported for Qwen3.5. Keep the final artifact easy to locate after the reasoning, such as a delimited conclusion, evidence table or separate code proposal.
The card recommends retaining final answers without thinking content in historical assistant messages. That differs from historical reasoning-preservation options and defaults introduced in later releases. Follow this exact checkpoint’s template and keep corrections or tool observations associated with the task they update.
Agentic usage
Qwen documents tool calling and recommends Qwen-Agent with configured or MCP-connected tools, and Qwen Code for coding-agent workflows. State the permitted operations, available evidence and observable completion criteria. An application executes tools and supplies their results; the model can use those observations to revise its plan or explanation. A proposed command or patch remains distinct from an action actually completed in the environment.
Ultra-long text and video
The maker’s extension uses YaRN with the checkpoint’s own interleaved positional settings. A factor of four targets the roughly one-million-token workflow; a factor of two is suggested for 524,288-token workloads. Static scaling can affect shorter-text performance, so it should follow a real need for larger context. Organize a long input by document, page or code identifier and leave room for the result.
The supported video example uses two frames per second by default. For hour-scale material, the maker recommends a larger processor budget corresponding to roughly 224K video tokens for higher-rate sampling. That is a configuration-dependent recipe, not a universal media-duration, complete-recall or timestamp-precision guarantee.
Generation guidance and substantive limits
This checkpoint’s Best Practices recommend temperature 0.6, top_p 0.95, top_k 20 and min_p 0 for thinking, and temperature 0.7, top_p 0.8, top_k 20 and min_p 0 for direct responses. These exact recommendations differ from some smaller Qwen3.5 editions. Where supported, a presence penalty between zero and two can reduce repetition, but higher values may mix languages or slightly reduce quality.
Recommended generation allowances are 32,768 tokens for most queries and 81,920 for unusually complex mathematical or programming work. They are recommendations, not independent hard output ceilings. Specify the required answer format and check calculations, extracted details and proposed code against the source. The useful output is a concrete textual artifact whose evidence and assumptions can be reviewed.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits qwen/qwen3.5-397b-a17b | 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 $0.00078 / 1K tokens Output · Output tokens $0.00468 / 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 $0.00078 / 1K tokens Output · Output tokens $0.00468 / 1K tokens Input · Cache read $0 / 1K tokens Input · Cache write $0 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "qwen/qwen3.5-397b-a17b",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${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': 'qwen/qwen3.5-397b-a17b', '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': 'qwen/qwen3.5-397b-a17b', '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 qwen/qwen3.5-397b-a17b
Use the method paths shown above.
Technical example in English
Reconcile a specification and implementation plan\n\nCompare spec S1: users may delete only their own drafts, and published records cannot be deleted. Plan P1: show Delete for every record owned by the user, regardless of status. Return the exact mismatch, a corrected eligibility rule and test cases for an owned draft, owned published record and another user’s draft.An evidence-linked mismatch, a corrected rule and three acceptance-test descriptions.Check that both ownership and draft status are required rather than treating either condition as sufficient.