# Qwen3.8-Flash-Next: official model and deployment card

Publisher model card with serving links and distinct thinking-mode settings. Use its exact artifact identity alongside the selected engine recipe.

- Type: Official model card
- Evidence: source_reported; site reproduced: false
- Model family: Qwen
- Model: Qwen/Qwen3.8-Flash-Next
- Original author: Qwen
- Original project: https://huggingface.co/Qwen/Qwen3.8-Flash-Next
- Pinned document: https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md
- Source revision: de4b8e4d43b917e7706784d8bb445c9af86a3540
- Source commit time (not experiment date): 2026-08-27T05:03:36.000Z
- Checked: 2026-10-17
- Experiment date: None
- Runtime topics: vLLM, SGLang, TokenSpeed
- Hardware topics: 
- Reported setup: {"system": null, "node_count": null, "parallelism": null, "kv_dtype": null, "context_limit_tokens": null, "experiment_date": null}
- Performance figures: attributed selections with per-point conditions; no cross-source ranking

## Conditions

Unknown is null. Recipe options are not measured benchmark settings.

```json
{
  "conditions": {
    "model_revision": "de4b8e4d43b917e7706784d8bb445c9af86a3540",
    "weight_hash": null,
    "runtime_execution_revision": null,
    "input_tokens": null,
    "output_tokens": null,
    "concurrency": null,
    "cache_state": null,
    "quantization": null,
    "speculative_decoding": null,
    "temperature": null,
    "top_p": null
  },
  "test_profiles": [],
  "recipe_parameters": null,
  "deployment_options": []
}
```

## Reading notes

- The card separates thinking and non-thinking sampling.
- Serving references point to model-specific vLLM, SGLang and TokenSpeed documentation.

## Limits

- Runtime support and usable context depend on the selected build and precision.

## License declaration

{"declared_spdx": "other", "declared_name": "qwen-community-1.0", "scope": "Publisher model-card declaration; inspect the linked LICENSE and artifact-specific terms."}


## Editorial answer

Publisher model card with serving links and distinct thinking-mode settings. Use its exact artifact identity alongside the selected engine recipe.

By Run AI Here editorial · 2026-10-17

AI-assisted source reading and original editorial synthesis; no independent model execution.

## How is the model’s size counted?

The card describes a 125B language-model component with 6B active parameters, plus 51B of n-gram embeddings and a 4B multi-token prediction component. The vLLM recipe uses a 176B summary label. Those counts cover different components; neither label should be substituted for the size of a downloaded artifact. The card lists 48 layers and a repeated hybrid of Gated DeltaNet and sparse attention.

Sources: https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L46

## Is the open checkpoint natively a million-token model?

The published native window is 262,144 tokens. Reaching 1,000,000 tokens requires the documented extension configuration; the hosted Qwen3.8-Flash product and the open Flash-Next checkpoint also have different production features. Record the actual RoPE settings and total input-plus-output budget before evaluating long tasks.

Sources: https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L71

## How do thinking and non-thinking settings differ?

Thinking mode uses temperature 1.0, top_p 0.95 and presence_penalty 0.0; non-thinking uses 0.7, 0.80 and 1.5 respectively. Both recommendations retain top_k 20, min_p 0.0 and repetition_penalty 1.0. The card also preserves historical thinking blocks by default. Changing that conversation policy changes both the prompt and cache behavior.

Sources: https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L342

## How should its benchmark table be read?

The coding evaluations specify harnesses, sampling and context. In particular, DeepSWE reports the better result from Claude Code and mini-SWE-agent rather than one universally shared harness. For vision tables, pass@3 and the mean over three attempts are different statistics. Keep the footnote for the selected task instead of treating every published cell as directly comparable.

Sources: https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L227

## Attributed configuration facts

```json
[
  {
    "name": "LM component / active / n-gram / MTP",
    "value": "125B / 6B / 51B / 4B",
    "classification": "publisher_declaration",
    "evidence_url": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L46",
    "source_document_sha256": "35ca37ccc366f1ba478dab33841a2c0c18ce53fd62f291ca05341f7728b225b2"
  },
  {
    "name": "Layers",
    "value": 48,
    "classification": "publisher_declaration",
    "evidence_url": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L50",
    "source_document_sha256": "35ca37ccc366f1ba478dab33841a2c0c18ce53fd62f291ca05341f7728b225b2"
  },
  {
    "name": "Native / extended context",
    "value": "262144 / up to 1000000 tokens",
    "classification": "publisher_declaration",
    "evidence_url": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L71",
    "source_document_sha256": "35ca37ccc366f1ba478dab33841a2c0c18ce53fd62f291ca05341f7728b225b2"
  },
  {
    "name": "Thinking sampling",
    "value": "temperature 1.0; top_p 0.95; top_k 20",
    "classification": "publisher_declaration",
    "evidence_url": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L342",
    "source_document_sha256": "35ca37ccc366f1ba478dab33841a2c0c18ce53fd62f291ca05341f7728b225b2"
  },
  {
    "name": "Non-thinking sampling",
    "value": "temperature 0.7; top_p 0.80; presence_penalty 1.5",
    "classification": "publisher_declaration",
    "evidence_url": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/de4b8e4d43b917e7706784d8bb445c9af86a3540/README.md#L343",
    "source_document_sha256": "35ca37ccc366f1ba478dab33841a2c0c18ce53fd62f291ca05341f7728b225b2"
  }
]
```