Alibaba / Qwen, released Feb 23, 2026
Qwen3.5 35B-A3Bprice, context, benchmarks and release details
Provisional: not enough results to rank yet Open weights 64% confidence 64 percent, Medium confidence, 2 of 7 expected sources in
62.7
SI Score
#49 of 142 ranked models
- Input, per 1M tokens
- $0.06Alibaba Model Studio pricingOfficial Alibaba Model Studio Global USD on-demand API; 0<Token≤128K; . Output uses non-thinking rate when both modes are available; thinking-only products use their thinking rate. Cache, Batch, free quotas and regional rates excluded
Retrieved Oct 9, 2026 · factual citation
Open source ↗ - Output, per 1M tokens
- $0.46Alibaba Model Studio pricingOfficial Alibaba Model Studio Global USD on-demand API; 0<Token≤128K; . Output uses non-thinking rate when both modes are available; thinking-only products use their thinking rate. Cache, Batch, free quotas and regional rates excluded
Retrieved Oct 9, 2026 · factual citation
Open source ↗ - Context window
- 262Kmodels.devPublished source fact
Retrieved Oct 9, 2026 · MIT
Open source ↗ - Max output
- 65.5Kmodels.devPublished source fact
Retrieved Oct 9, 2026 · MIT
Open source ↗ - Released
- Feb 23, 2026models.devPublished source fact
Retrieved Oct 9, 2026 · MIT
Open source ↗
How the score breaks down
Coding (weight 40 percent) —
Math (weight 15 percent) 62.2
Preference (weight 15 percent) 72.6
Reasoning (weight 30 percent) 82.3
Weights: reasoning 30%, math 15%, coding 40%, preference 15%. Results use fixed 0–100 scales before averaging, and thin evidence is pulled toward 50. Method si-v3-retained-evidence-2, computed Oct 9, 2026, 05:13 UTC.
Around it on the leaderboard
- 47 Muse Spark 1.1 62.9
- 48 Qwen3.7 Plus 62.7
- 49 Qwen3.5 35B-A3B 62.7
- 50 Claude Opus 4.5 62.6
- 51 MiMo-V2.5-Pro 62.6
Benchmark results
3 benchmarks, 5 resultsEach row shows the best published result. Where a model was tested at several settings, such as reasoning effort, open the row to see each one. Hover or tap a value for its source.
GPQA Diamondreasoning 83.5%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] Published Aug 28, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗
83.5 2 settings
- no reasoning 81.2%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] nonePublished Aug 28, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗ - Setting 2 83.5%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] Published Aug 28, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗
OTIS Mock AIME 2024–2025math 70%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] nonePublished Aug 30, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗
70.0 2 settings
- no reasoning 70%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] nonePublished Aug 30, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗ - Setting 2 54.4%Epoch AI BenchmarkingEpoch-owned evaluation mean score; scores only, no benchmark questions [variant] Published Aug 30, 2026
Retrieved Oct 9, 2026 · CC-BY
Open source ↗
LMArena Textpreference 1394.5 eloLMArena / ArenaPublished source fact [variant] text / overallPublished Oct 8, 2026
Retrieved Oct 9, 2026 · CC-BY-4.0
Open source ↗ 72.6
Open source ↗ 72.6
Normalization uses a fixed 0–100 scale for each unit, independent of other models. Compare evaluation conditions before reading a small gap as decisive. “Lab-reported” marks the provider's own published figure.
Details and sources
- Open weights
- Yesmodels.devPublished source fact
Retrieved Oct 9, 2026 · MIT
Open source ↗ - License
- not yet reported
- Input modalities
- text, image, video, audiomodels.devPublished source fact
Retrieved Oct 9, 2026 · MIT
Open source ↗ - First seen by SuperIndex
- Oct 8, 2026
- Coverage
- 51% of expected source weight
Reported (2)
- Epoch AI BenchmarkingOct 8, 2026
- LMArena / ArenaOct 8, 2026
Awaiting (5)
- ARC Prize13% of weight
- Humanity’s Last Exam13% of weight
- Official model cards via models.dev4% of weight
- LiveBench13% of weight
- Terminal-Bench6% of weight
Confidence rises as pending sources publish. Some sources never cover some models, so confidence reaches 100% at 80% of expected weight.