Qwen3.8 highlights: sustained text reasoning
Qwen3.8-2.4T-A95B is the open Qwen3.8 checkpoint for text reasoning. Qwen describes improvements across coding, professional work, research and long-horizon agent tasks, including planning and the use of environment feedback. It accepts text and returns text. The maker explicitly separates it from the managed Qwen3.8-Max offering: vision input and non-thinking support are additional features of that offering, not capabilities of this exact open checkpoint.
For a coding task, provide the relevant modules, the requirement and an input that exposes the current behavior. Request an explained change and a final artifact that can be inspected. For a research or professional report, provide the evidence, required sections and evaluation criteria, then ask the model to distinguish supported findings from assumptions. These are practical applications of the documented task scope, not promises of executed code or expert-certified conclusions.
Model overview
The sparse mixture-of-experts model has 2.4 trillion total parameters and activates 95 billion. It has 92 layers and 512 routed experts, selecting ten routed experts together with one shared expert. Its hybrid blocks combine Gated DeltaNet and gated attention with expert layers, following a three-to-one pattern. Multi-token prediction is part of its training. The activated parameter count is not the full checkpoint size, and a large total count does not establish a response-time guarantee.
The native context is 262,144 tokens, with documented extension up to 1,010,000. This differs from the managed Max reference’s default one-million-token context. Use the actual configured sequence allowance for source text, conversation history, reasoning and the final response rather than assume all of the largest figures apply simultaneously.
Using the exact thinking checkpoint
Thinking is required for every interaction and cannot be disabled. The documented output begins with a reasoning block before the final text. reasoning_effort supports xhigh, medium and low; xhigh is the default. Low reduces the intended depth but remains a thinking mode. A request for a concise final answer can control the desired deliverable without converting this model into a non-thinking edition.
preserve_thinking is enabled by default. Retaining reasoning from earlier turns helps an iterative task carry its decision context forward. Keep corrections, tool results and the final answer associated with the turns that produced them. This model should follow its own history contract rather than inherit older advice to remove every reasoning block from past messages. The exact exposure of effort and preservation controls depends on the interface implementing the checkpoint recipe.
Harnesses and environment feedback
Qwen describes compatibility with development harnesses and tools. A sustained task can be expressed as a goal, permitted actions and evidence for completion; the application can then provide observations from inspection or execution. Ask the model to explain how each new result changes the plan and to report what remains unresolved. Planning an action and receiving a real result are different parts of that workflow.
The card’s managed Max examples include additional built-in tools. They should not be converted into automatic web search, code execution or visual input for this open model. Its useful role is a text reasoning component in a connected workflow; tools and external effects belong to the application supplying them.
Generation recommendations
The maker recommends temperature 1.0, top_p 0.95, top_k 20, min_p 0, presence penalty 0 and repetition penalty 1.0. Where supported, a presence penalty between zero and two can reduce repeated text, but higher values may mix languages or slightly reduce quality. Parameter support varies with the implementation.
For a configured one-million-token workflow with separate reasoning and final-response limits, Qwen recommends 262,144 tokens for reasoning and 131,072 for the final answer. Those are conditional allowances, not separately established hard ceilings or budgets that fit together within the native 262,144-token context. Allocate enough space for a complete artifact and ask for a final result that is easy to find after the reasoning.
Practical limits
This checkpoint does not accept images, audio or video. Convert relevant evidence into a suitable text representation through a separate workflow when needed, retaining identifiers that allow the answer to be checked. The card states an extended context but does not give a complete per-task extension recipe or a universal media limit for this text-only model. Preserve the exact checkpoint identity and configuration, check calculations and code proposals against the source, and treat the final artifact as a result to review rather than proof that a whole external task has been completed.
", "aboutText": "Qwen3.8 highlights: sustained text reasoning\n\nQwen3.8-2.4T-A95B is the open Qwen3.8 checkpoint for text reasoning. Qwen describes improvements across coding, professional work, research and long-horizon agent tasks, including planning and the use of environment feedback. It accepts text and returns text. The maker explicitly separates it from the managed Qwen3.8-Max offering: vision input and non-thinking support are additional features of that offering, not capabilities of this exact open checkpoint.\n\nFor a coding task, provide the relevant modules, the requirement and an input that exposes the current behavior. Request an explained change and a final artifact that can be inspected. For a research or professional report, provide the evidence, required sections and evaluation criteria, then ask the model to distinguish supported findings from assumptions. These are practical applications of the documented task scope, not promises of executed code or expert-certified conclusions.\n\nModel overview\n\nThe sparse mixture-of-experts model has 2.4 trillion total parameters and activates 95 billion. It has 92 layers and 512 routed experts, selecting ten routed experts together with one shared expert. Its hybrid blocks combine Gated DeltaNet and gated attention with expert layers, following a three-to-one pattern. Multi-token prediction is part of its training. The activated parameter count is not the full checkpoint size, and a large total count does not establish a response-time guarantee.\n\nThe native context is 262,144 tokens, with documented extension up to 1,010,000. This differs from the managed Max reference’s default one-million-token context. Use the actual configured sequence allowance for source text, conversation history, reasoning and the final response rather than assume all of the largest figures apply simultaneously.\n\nUsing the exact thinking checkpoint\n\nThinking is required for every interaction and cannot be disabled. The documented output begins with a reasoning block before the final text. reasoning_effort supports xhigh, medium and low; xhigh is the default. Low reduces the intended depth but remains a thinking mode. A request for a concise final answer can control the desired deliverable without converting this model into a non-thinking edition.\n\npreserve_thinking is enabled by default. Retaining reasoning from earlier turns helps an iterative task carry its decision context forward. Keep corrections, tool results and the final answer associated with the turns that produced them. This model should follow its own history contract rather than inherit older advice to remove every reasoning block from past messages. The exact exposure of effort and preservation controls depends on the interface implementing the checkpoint recipe.\n\nHarnesses and environment feedback\n\nQwen describes compatibility with development harnesses and tools. A sustained task can be expressed as a goal, permitted actions and evidence for completion; the application can then provide observations from inspection or execution. Ask the model to explain how each new result changes the plan and to report what remains unresolved. Planning an action and receiving a real result are different parts of that workflow.\n\nThe card’s managed Max examples include additional built-in tools. They should not be converted into automatic web search, code execution or visual input for this open model. Its useful role is a text reasoning component in a connected workflow; tools and external effects belong to the application supplying them.\n\nGeneration recommendations\n\nThe maker recommends temperature 1.0, top_p 0.95, top_k 20, min_p 0, presence penalty 0 and repetition penalty 1.0. Where supported, a presence penalty between zero and two can reduce repeated text, but higher values may mix languages or slightly reduce quality. Parameter support varies with the implementation.\n\nFor a configured one-million-token workflow with separate reasoning and final-response limits, Qwen recommends 262,144 tokens for reasoning and 131,072 for the final answer. Those are conditional allowances, not separately established hard ceilings or budgets that fit together within the native 262,144-token context. Allocate enough space for a complete artifact and ask for a final result that is easy to find after the reasoning.\n\nPractical limits\n\nThis checkpoint does not accept images, audio or video. Convert relevant evidence into a suitable text representation through a separate workflow when needed, retaining identifiers that allow the answer to be checked. The card states an extended context but does not give a complete per-task extension recipe or a universal media limit for this text-only model. Preserve the exact checkpoint identity and configuration, check calculations and code proposals against the source, and treat the final artifact as a result to review rather than proof that a whole external task has been completed.", "editorialOverlay": { "copy_fields": { "practical_html": "Technical example in English
Review a resumable migration plan\n\nReview this migration design against the requirement that every row is copied exactly once. Step 1 copies 100 rows after last_id. Step 2 saves last_id. A crash can occur between steps. Explain the duplicate risk after restart and propose a minimal idempotent write or atomic checkpoint design. Include a test that interrupts precisely between the two steps.A cause analysis, one concrete correction strategy and a crash-point regression scenario.Rerunning after the interruption cannot produce duplicate destination rows or skip uncopied source rows.qwen/qwen3.8-2.4t-a95bQwen3.8-2.4T-A95B is Qwen’s sparse mixture-of-experts text reasoning model, with 2.4 trillion total parameters and 95 billion activated. This open checkpoint always uses thinking, supports adjustable reasoning effort and retains reasoning context for sustained coding and professional tasks.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Qwen3.8 highlights: sustained text reasoning
Qwen3.8-2.4T-A95B is the open Qwen3.8 checkpoint for text reasoning. Qwen describes improvements across coding, professional work, research and long-horizon agent tasks, including planning and the use of environment feedback. It accepts text and returns text. The maker explicitly separates it from the managed Qwen3.8-Max offering: vision input and non-thinking support are additional features of that offering, not capabilities of this exact open checkpoint.
For a coding task, provide the relevant modules, the requirement and an input that exposes the current behavior. Request an explained change and a final artifact that can be inspected. For a research or professional report, provide the evidence, required sections and evaluation criteria, then ask the model to distinguish supported findings from assumptions. These are practical applications of the documented task scope, not promises of executed code or expert-certified conclusions.
Model overview
The sparse mixture-of-experts model has 2.4 trillion total parameters and activates 95 billion. It has 92 layers and 512 routed experts, selecting ten routed experts together with one shared expert. Its hybrid blocks combine Gated DeltaNet and gated attention with expert layers, following a three-to-one pattern. Multi-token prediction is part of its training. The activated parameter count is not the full checkpoint size, and a large total count does not establish a response-time guarantee.
The native context is 262,144 tokens, with documented extension up to 1,010,000. This differs from the managed Max reference’s default one-million-token context. Use the actual configured sequence allowance for source text, conversation history, reasoning and the final response rather than assume all of the largest figures apply simultaneously.
Using the exact thinking checkpoint
Thinking is required for every interaction and cannot be disabled. The documented output begins with a reasoning block before the final text. reasoning_effort supports xhigh, medium and low; xhigh is the default. Low reduces the intended depth but remains a thinking mode. A request for a concise final answer can control the desired deliverable without converting this model into a non-thinking edition.
preserve_thinking is enabled by default. Retaining reasoning from earlier turns helps an iterative task carry its decision context forward. Keep corrections, tool results and the final answer associated with the turns that produced them. This model should follow its own history contract rather than inherit older advice to remove every reasoning block from past messages. The exact exposure of effort and preservation controls depends on the interface implementing the checkpoint recipe.
Harnesses and environment feedback
Qwen describes compatibility with development harnesses and tools. A sustained task can be expressed as a goal, permitted actions and evidence for completion; the application can then provide observations from inspection or execution. Ask the model to explain how each new result changes the plan and to report what remains unresolved. Planning an action and receiving a real result are different parts of that workflow.
The card’s managed Max examples include additional built-in tools. They should not be converted into automatic web search, code execution or visual input for this open model. Its useful role is a text reasoning component in a connected workflow; tools and external effects belong to the application supplying them.
Generation recommendations
The maker recommends temperature 1.0, top_p 0.95, top_k 20, min_p 0, presence penalty 0 and repetition penalty 1.0. Where supported, a presence penalty between zero and two can reduce repeated text, but higher values may mix languages or slightly reduce quality. Parameter support varies with the implementation.
For a configured one-million-token workflow with separate reasoning and final-response limits, Qwen recommends 262,144 tokens for reasoning and 131,072 for the final answer. Those are conditional allowances, not separately established hard ceilings or budgets that fit together within the native 262,144-token context. Allocate enough space for a complete artifact and ask for a final result that is easy to find after the reasoning.
Practical limits
This checkpoint does not accept images, audio or video. Convert relevant evidence into a suitable text representation through a separate workflow when needed, retaining identifiers that allow the answer to be checked. The card states an extended context but does not give a complete per-task extension recipe or a universal media limit for this text-only model. Preserve the exact checkpoint identity and configuration, check calculations and code proposals against the source, and treat the final artifact as a result to review rather than proof that a whole external task has been completed.
POST /v1/chat/completionstexttext| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits qwen/qwen3.8-2.4t-a95b | 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 $0.004 / 1K tokens Output · Output tokens $0.012 / 1K tokens Input · Cache read $0.0005 / 1K tokens Input · Cache write $0.005 / 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 $0.004 / 1K tokens Output · Output tokens $0.012 / 1K tokens Input · Cache read $0.0005 / 1K tokens Input · Cache write $0.005 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
set -eu\n: "${AILOOT_API_KEY:?Set your Ailoot API key}"\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads('{"model": "qwen/qwen3.8-2.4t-a95b", "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/qwen3.8-2.4t-a95b',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/qwen3.8-2.4t-a95b',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/qwen3.8-2.4t-a95b
Use the method paths shown above.
qwen/qwen3.8-2.4t-a95bQwen3.8-2.4T-A95B is Qwen’s sparse mixture-of-experts text reasoning model, with 2.4 trillion total parameters and 95 billion activated. This open checkpoint always uses thinking, supports adjustable reasoning effort and retains reasoning context for sustained coding and professional tasks.
This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.
Qwen3.8 highlights: sustained text reasoning
Qwen3.8-2.4T-A95B is the open Qwen3.8 checkpoint for text reasoning. Qwen describes improvements across coding, professional work, research and long-horizon agent tasks, including planning and the use of environment feedback. It accepts text and returns text. The maker explicitly separates it from the managed Qwen3.8-Max offering: vision input and non-thinking support are additional features of that offering, not capabilities of this exact open checkpoint.
For a coding task, provide the relevant modules, the requirement and an input that exposes the current behavior. Request an explained change and a final artifact that can be inspected. For a research or professional report, provide the evidence, required sections and evaluation criteria, then ask the model to distinguish supported findings from assumptions. These are practical applications of the documented task scope, not promises of executed code or expert-certified conclusions.
Model overview
The sparse mixture-of-experts model has 2.4 trillion total parameters and activates 95 billion. It has 92 layers and 512 routed experts, selecting ten routed experts together with one shared expert. Its hybrid blocks combine Gated DeltaNet and gated attention with expert layers, following a three-to-one pattern. Multi-token prediction is part of its training. The activated parameter count is not the full checkpoint size, and a large total count does not establish a response-time guarantee.
The native context is 262,144 tokens, with documented extension up to 1,010,000. This differs from the managed Max reference’s default one-million-token context. Use the actual configured sequence allowance for source text, conversation history, reasoning and the final response rather than assume all of the largest figures apply simultaneously.
Using the exact thinking checkpoint
Thinking is required for every interaction and cannot be disabled. The documented output begins with a reasoning block before the final text. reasoning_effort supports xhigh, medium and low; xhigh is the default. Low reduces the intended depth but remains a thinking mode. A request for a concise final answer can control the desired deliverable without converting this model into a non-thinking edition.
preserve_thinking is enabled by default. Retaining reasoning from earlier turns helps an iterative task carry its decision context forward. Keep corrections, tool results and the final answer associated with the turns that produced them. This model should follow its own history contract rather than inherit older advice to remove every reasoning block from past messages. The exact exposure of effort and preservation controls depends on the interface implementing the checkpoint recipe.
Harnesses and environment feedback
Qwen describes compatibility with development harnesses and tools. A sustained task can be expressed as a goal, permitted actions and evidence for completion; the application can then provide observations from inspection or execution. Ask the model to explain how each new result changes the plan and to report what remains unresolved. Planning an action and receiving a real result are different parts of that workflow.
The card’s managed Max examples include additional built-in tools. They should not be converted into automatic web search, code execution or visual input for this open model. Its useful role is a text reasoning component in a connected workflow; tools and external effects belong to the application supplying them.
Generation recommendations
The maker recommends temperature 1.0, top_p 0.95, top_k 20, min_p 0, presence penalty 0 and repetition penalty 1.0. Where supported, a presence penalty between zero and two can reduce repeated text, but higher values may mix languages or slightly reduce quality. Parameter support varies with the implementation.
For a configured one-million-token workflow with separate reasoning and final-response limits, Qwen recommends 262,144 tokens for reasoning and 131,072 for the final answer. Those are conditional allowances, not separately established hard ceilings or budgets that fit together within the native 262,144-token context. Allocate enough space for a complete artifact and ask for a final result that is easy to find after the reasoning.
Practical limits
This checkpoint does not accept images, audio or video. Convert relevant evidence into a suitable text representation through a separate workflow when needed, retaining identifiers that allow the answer to be checked. The card states an extended context but does not give a complete per-task extension recipe or a universal media limit for this text-only model. Preserve the exact checkpoint identity and configuration, check calculations and code proposals against the source, and treat the final artifact as a result to review rather than proof that a whole external task has been completed.
POST /v1/messagestextjson, sse| Field | Type | Requirement | Values & limits | Default |
|---|---|---|---|---|
Fieldmodel | Typestring | RequirementRequired | Values & limits qwen/qwen3.8-2.4t-a95b | 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 $0.004 / 1K tokens Output · Output tokens $0.012 / 1K tokens Input · Cache read $0.0005 / 1K tokens Input · Cache write $0.005 / 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 $0.004 / 1K tokens Output · Output tokens $0.012 / 1K tokens Input · Cache read $0.0005 / 1K tokens Input · Cache write $0.005 / 1K tokens Input · Cache write, 1 hour — / 1K tokens |
Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →
{\n "model": "qwen/qwen3.8-2.4t-a95b",\n "max_tokens": 128,\n "messages": [\n {\n "role": "user",\n "content": "Summarize the incident report."\n }\n ],\n "stream": false\n}set -eu\n: "${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/qwen3.8-2.4t-a95b', '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/qwen3.8-2.4t-a95b', '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/qwen3.8-2.4t-a95b
Use the method paths shown above.
Technical example in English
Review a resumable migration plan\n\nReview this migration design against the requirement that every row is copied exactly once. Step 1 copies 100 rows after last_id. Step 2 saves last_id. A crash can occur between steps. Explain the duplicate risk after restart and propose a minimal idempotent write or atomic checkpoint design. Include a test that interrupts precisely between the two steps.A cause analysis, one concrete correction strategy and a crash-point regression scenario.Rerunning after the interruption cannot produce duplicate destination rows or skip uncopied source rows.