BlogLLM Benchmark Scores as Published, Side by Side
Qwen3.8-2.4T-A95B: Benchmark Scores as Published
Every result the catalog holds for Qwen3.8-2.4T-A95B: 19 tests in 8 areas, from 3 pages. Each was printed by the company named beside it. Each number is as printed; nothing is ranked or rescored.
Developer AlibabaContext window 262KMax output 262KInput TextOutput TextWeights OpenData version 2026103001, as of 2026-10-30
Knowledge Work
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AutomationBench | 57.2% | Mistral AI, Mistral Large 4 (preview)checked 2026-10-15 | - |
| GDPval-AA v2 | 1613 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| JobBench | 52.7 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
Agentic Coding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Terminal-Bench 2.1 | 82% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh; Harness: Terminus 2 |
| DeepSWE v1.1 | 57% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh; Harness: preferred harness of each model |
| SWE-Atlas-QnA (setting not stated) | 62% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh; Harness: preferred harness of each model |
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPQA Diamond | 93.5% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| SciCode | 54.1% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| SciCode-Verified | 93.8% | Mistral AI, Mistral Large 4 (preview)checked 2026-10-15 | Attempt budget: pass@1 (n=6) |
Cybersecurity
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Artificial Analysis Cyber Index (successes) | 13 | Mistral AI, Mistral Large 4 (preview)checked 2026-10-15 | - |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Humanity's Last Exam (with tools) | 56.2% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh; Tools: with tools |
| AA-Omniscience Index | 4 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| Humanity's Last Exam Diamond (no tools) | 17.3% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh; Tools: none |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Toolathlon Verified | 72.5% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| MCP Atlas | 84.5% | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
Safety and Alignment
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| FORTRESS adversarial (static), ARSlower is better | 21.6 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| FORTRESS adversarial (static), ORSlower is better | 12.8 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
| StrongREJECT v2 (adaptive)lower is better | 65.6 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
Preference and Composite Rankings
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Artificial Analysis Intelligence Index | 40 | Meta, Meta Muse Spark Open 816B-A42B model cardchecked 2026-10-19 | Effort: xhigh |
Price
API list prices in US dollars per one million tokens, as printed on the developer's pricing page on the day checked.
| Price | Value | Printed by | Setup as printed |
|---|---|---|---|
| Input | $1.65 | Alibaba, Alibaba Model Studio pricingchecked 2026-10-15 | - |
| Output | $4.951 | Alibaba, Alibaba Model Studio pricingchecked 2026-10-15 | - |
More from Alibaba
- Qwen3.8-Max
- Qwen3.7-Plus
- Qwen3.6-27B
- Qwen3.6-35B-A3B
- Qwen3.6-Plus
- Qwen3.5-122B-A10B
- Qwen3.5-27B
- Qwen3.5-35B-A3B
- Qwen3.5-397B-A17B
- Qwen3-VL-32B-Instruct
- Qwen3-VL-32B-Thinking
- Qwen3-VL-8B-Instruct
- Qwen3-VL-8B-Thinking
- Qwen3-VL-235B-A22B
- Qwen3-VL-235B-A22B-Thinking
- Qwen3-Next-80B-A3B
- Qwen3-30B-A3B-Instruct-2507
- Qwen3-30B-A3B-Thinking-2507
- Qwen3-Coder-Plus
- Qwen3-235B-A22B
- Qwen3-32B
- Qwen3-8B
- QwQ-32B-Preview
- Qwen-Max
- Qwen3-Max-Thinking
- Qwen3.7-Max
- Qwen3.8-27B
- Qwen3.8-Flash-Next
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.