AlibabaQwen3-Next 80B A3B
Qwen3-Next 80B A3B is an Alibaba open-weights reasoning model, released 11 Sept 2025. Its best configuration (thinking reasoning effort) ranks #401 of 785 on the BenchLeader Index at 49.3 ±4.3, in the lower half. The ± is the point: 309 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in maths (57) and lowest in agents & tools (37). At $0.350 per million tokens blended it is cheaper than most ranked models. Output speed of 109 tokens per second puts it faster than most, with a first answer in 0.8 s. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Oct 2026.
- Blended price
- $0.350/M
- $0.100 in · $1.10 out
- Output speed
- 109 tok/s
- measured by Artificial Analysis
- First answer
- 0.78 s
- first token 0.78 s
- Context
- 131k
- Full answer
- 5.39 s
- median, reasoning included
- Cost per run
- –
- Released
- 11 Sept 2025
- Overall index49
- Reasoning49
- Coding55
- Agents & tools37
- Maths57
- Knowledge38
- Instruction following55
- Human preference54
- Long context57
- Composite42
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| thinkingbest | 49.3 | #401 | 109 tok/s | 0.78 s | $0.0004 | Agents & tools 37 · Coding 55 · Composite 42 · Human preference 54 · Instruction following 55 · Knowledge 38 · Long context 57 · Maths 57 · Reasoning 49 |
| not stated | 48.3 | #431 | 109 tok/s | 0.78 s | $0.0004 | Agents & tools 40 · Coding 57 · Composite 40 · Human preference 57 · Instruction following 42 · Knowledge 35 · Long context 53 · Maths 59 · Reasoning 48 |
How its index has moved
9 Sept 2026 to 21 Oct 2026- thinking
- not stated
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Arena Hard Prompts | 1385#181 | 1419#148 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 75.9%#244 | 73.8%#263 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 12.6%#278 | 7.7%#358 | Artificial Analysis | |
| Kagi LLM Benchmark | 66.7%#36 | 54.4%#69 | Kagi LLM Benchmark | |
| GPQA Diamond (HELM)not in index | 63.0%#23 | – | HELM Capabilities | |
| CritPt | 0.0%#315 | 0.0%#315 | Artificial Analysis |
Coding
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Arena Coding | 1421#180 | 1444#150 | Arena (LMArena) | |
| Terminal-Bench 2.1 (AA)not in index | 6.7%#199 | – | Artificial Analysis |
Agents & tools
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Terminal-Bench Hard | 9.8%#228 | 7.6%#241 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 41.5%#219 | 21.6%#319 | Artificial Analysis | |
| GDPval-AA v2.1not in index | 0.0%#124 | – | Artificial Analysis | |
| τ³-Banking (AA) | 6.2%#169 | – | Artificial Analysis |
Maths
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Omni-MATH (HELM) | 46.7%#22 | – | HELM Capabilities | |
| Arena Maths | 1392#159 | 1414#129 | Arena (LMArena) |
Knowledge
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| AA-Omniscience | -51.4#400 | -59.5#463 | Artificial Analysis | |
| MMLU-Pro (HELM)not in index | 78.6%#23 | – | HELM Capabilities | |
| AA-Omniscience: accuracynot in index | 19.0%#321 | 17.3%#369 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 13.0%#365 | 7.3%#486 | Artificial Analysis |
Instruction following
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| IFBench | 60.7%#125 | 39.7%#264 | Artificial Analysis | |
| IFEval (HELM) | 81.0%#44 | – | HELM Capabilities | |
| WildBench (HELM) | 80.7%#29 | – | HELM Capabilities | |
| Arena Instruction Followingnot in index | 1358#183 | 1376#166 | Arena (LMArena) |
Human preference
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Arena Text | 1369#183 | 1399#151 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1326#190 | 1313#202 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1349#195 | 1401#152 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1370#187 | 1389#171 | Arena (LMArena) |
Long context
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Fiction.LiveBench 120keffort not stated | – | 62.5%#13 | Fiction.live | |
| AA-LCR | 63.7%#264 | 52.7%#314 | Artificial Analysis |
Composite
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| AA Intelligence Index v4.3.2 | 11.2#359 | 9.6#392 | Artificial Analysis | |
| HELM Capabilities meannot in index | 70.0%#23 | – | HELM Capabilities |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Parasail | 118 tok/s | 0.45 s | $0.100 | $1.10 | 262k | fp8 |
| DeepInfra | 114 tok/s | 0.56 s | $0.090 | $1.10 | 262k | fp8 |
| DeepInfra | 114 tok/s | 0.47 s | $0.135 | $1.65 | 262k | fp8 |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0004 | – | 3.5 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0019 | – | 6.3 s |
| Code edit | 6,000 / 1,500 | $0.0023 | – | 14.6 s |
| Agentic coding session | 60,000 / 4,000 | $0.010 | – | 37.6 s |
| Structured extraction | 2,000 / 200 | $0.0004 | – | 2.6 s |
See also
Questions people ask
- What are Qwen3-Next 80B A3B's benchmark scores?
- As of 24 Oct 2026, Qwen3-Next 80B A3B (thinking effort) scores 49.3 ±4.3 on the BenchLeader Index, strongest in maths (57) and weakest in agents & tools (37). That rests on 15 independent benchmark results, and every score on this page links to its publisher.
- How does Qwen3-Next 80B A3B rank?
- Qwen3-Next 80B A3B (thinking effort) is #401 of 785 ranked configurations on 24 Oct 2026. 309 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Qwen3-Next 80B A3B cost per 1M tokens?
- As of 24 Oct 2026, Qwen3-Next 80B A3B costs $0.100 per 1M input tokens and $1.10 per 1M output tokens, $0.350 blended (three input tokens per output token). The list price last changed on 24 Oct 2026.
- When was Qwen3-Next 80B A3B released?
- Alibaba released Qwen3-Next 80B A3B on 11 Sept 2025. BenchLeader has results for it at 2 reasoning-effort settings.
Data as of 24 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Qwen3-Next 80B A3B: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/qwen3-next-80b-a3b.