Model comparison
Qwen3.8-2.4T-A95B vs Qwen3.8-Max
Two Qwen tiers compared on the figures that decide which one a workload actually needs.
Catalog record checked August 14, 2026; individual provider fields may change.
Qwen
Qwen3.8-2.4T-A95B
Frontier · Open weights
Qwen
Qwen3.8-Max
Frontier
| Specification | Qwen3.8-2.4T-A95B | Qwen3.8-Max |
|---|---|---|
| Provider | Qwen | Qwen |
| Tier | Frontier | Frontier |
| Context window | 262K | Winner: 1M |
| Max output | 131K | 131K |
| Input / 1M tokens | Not verified | $2 |
| Output / 1M tokens | Not verified | $6 |
| Weights | Open | Closed |
| Parameters | 2.4T total / 95B active (MoE) | 2.4T total / ~95B active (MoE) |
| Reasoning levels | low, medium, xhigh | Not verified |
| Modalities | text | text, image, video |
| Released | August 12, 2026 | August 3, 2026 |
| Artificial Analysis Intelligence Index (2026-08-08) | Not verified | 58 |
Prices are USD per million tokens at standard rates, excluding batch and caching discounts. Bold indicates the better figure where one is objectively better. Values we could not confirm from the provider are shown as “Not verified” rather than estimated.
Qwen3.8-2.4T-A95B
Official Hugging Face checkpoint for Qwen3.8 open weights — a text base model, not the hosted Qwen3.8-Max API.
Best for
- Self-hosting the Qwen3.8 flagship
- vLLM / SGLang serving
- Evaluating the open Max-class weights
Watch out
Native context is 262,144 tokens (YaRN extends toward ~1,010,000). Hugging Face best-practice caps: 262,144 reasoning tokens and 131,072 final-response tokens. Thinking cannot be disabled. Qwen3.8-Max License: MaaS / AI Work Assistant businesses over $50M trailing 12-month revenue need a separate commercial licence. Self-hosting still needs datacentre-class memory. Do not treat this checkpoint as the multimodal 1M-context API.
Qwen3.8-Max
Alibaba's hosted Max-class API — 2.4T MoE with a 1M context window (max input 991,808), vision, and managed tools. Distinct from the Hugging Face Qwen3.8-2.4T-A95B checkpoint.
Best for
- Agentic coding
- Multimodal knowledge work
- Long-context delivery
Watch out
This row is the hosted Qwen3.8-Max API, not the downloadable checkpoint. The open-weight Qwen3.8-2.4T-A95B card is catalogued separately and does not include the API's vision, non-thinking, or built-in tool extras.
Benchmark
DeepSWE 1.1 in context
Only Qwen3.8-Max has a published DeepSWE 1.1 result. It is shown against the wider field, with cost per completed task alongside the score.
Local leader
Claude Opus 5 [max]
74%
Rows shown
24
Highest published reasoning effort per model (not best Pass@1)
Snapshot date
2026-08-13
Mirrored from deepswe.datacurve.ai
Better is toward the top-right (higher pass rate, lower cost). X-axis is reversed to match DeepSWE’s public chart. v1.1 uses average cost / tokens / steps; v1 uses published medians.
| # | Model | Pass@1 | Cost / task | Tokens / task | Steps / task |
|---|---|---|---|---|---|
| 1 | Claude Opus 5 [max] | 74% | $11.84 | 118k | 99 |
| 2 | GPT-5.6 Sol [max] | 73% | $8.39 | 60k | 61 |
| 3 | Claude Fable 5 [max] | 70% | $21.63 | 119k | 88 |
| 4 | GPT-5.6 Terra [max] | 70% | $4.95 | 72k | 76 |
| 5 | Kimi K3 [max] | 69% | $4.65 | 82k | 98 |
| 6 | GPT-5.6 Luna [max] | 67% | $3.03 | 73k | 102 |
| 7 | GPT-5.5 [xhigh] | 67% | $7.23 | 46k | 82 |
| 8 | Grok 4.6 [xhigh] | 67% | $5.50 | 71k | 87 |
| 9 | Gemini 3.7 Flash [high] | 65% | $2.18 | 107k | 125 |
| 10 | DeepSeek V4-Pro [max] | 63% | $0.24 | 106k | 155 |
| 11 | Claude Opus 4.8 [max] | 59% | $13.22 | 135k | 120 |
| 12 | Qwen3.8-Max [xhigh] | 58% | $3.73 | 95k | 111 |
| 13 | Muse Spark 1.2 [xhigh] | 55% | $3.70 | 99k | 101 |
| 14 | Claude Sonnet 5 [max] | 54% | $26.40 | 214k | 268 |
| 15 | Grok 4.5 [high] | 54% | $2.42 | 36k | 61 |
| 16 | DeepSeek V4-Flash [max] | 53% | $0.10 | 108k | 153 |
| 17 | Muse Spark 1.1 [xhigh] | 53% | $2.36 | 74k | 96 |
| 18 | GPT-5.4 [xhigh] | 52% | $5.65 | 71k | 70 |
| 19 | Gemini 3.6 Flash [high] | 47% | $4.42 | 96k | 117 |
| 20 | GLM 5.2 [max] | 44% | $3.92 | 78k | 129 |
| 21 | Gemini 3.5 Flash [high] | 36% | $3.45 | 76k | 105 |
| 22 | Kimi K2.7 Code | 31% | $2.82 | 59k | 149 |
| 23 | Claude Sonnet 4.6 [high] | 30% | $5.52 | 76k | 134 |
| 24 | Gemini 3.1 Pro [high] | 12% | $2.14 | 28k | 76 |
DeepSWE “Best” picks the highest published reasoning effort per model (not the highest pass rate). Small gaps may not be statistically meaningful — confirm on deepswe.datacurve.ai.
Common questions
Qwen3.8-2.4T-A95B vs Qwen3.8-Max
Answered from the verified figures on this page rather than general guidance.
Which has the larger context window, Qwen3.8-2.4T-A95B or Qwen3.8-Max?
Qwen3.8-Max accepts 1M tokens against 262K for Qwen3.8-2.4T-A95B. This only matters if you routinely send very long documents or large codebases.
Should I use Qwen3.8-2.4T-A95B or Qwen3.8-Max?
Both sit in the frontier tier, so the choice usually comes down to price and context rather than capability. Qwen3.8-2.4T-A95B suits self-hosting the qwen3.8 flagship; Qwen3.8-Max suits agentic coding.
Can I self-host Qwen3.8-2.4T-A95B or Qwen3.8-Max?
Qwen3.8-2.4T-A95B publishes open weights, but self-hosting and third-party provider access depend on the licence, hardware, serving support, and availability. Qwen3.8-Max is a closed model whose supported access paths are controlled by its provider.
Next step
Choosing between them
Tier and workload decide this more reliably than a leaderboard position does.
If both sit in the same tier, the decision usually comes down to context window and output price rather than headline capability — output tokens dominate real bills.
If one is a step up within the same provider, the useful question is whether your hardest task actually fails on the cheaper tier. Most production volume — classification, extraction, summarization — does not.
Choosing a harness rather than a model — Cursor, Copilot, Claude Code, Muse Code, or Lovable? Compare agentic harnesses · Latest releases.