math benchmark
SciCode Leaderboard
SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.
Updated Aug 17, 2026
Models21
Model coverage21
MetricScore
EvidenceB
SciCode Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelBY | Score59.8% | Percentile100.0% | Participants21 | EvidenceC | Evaluated |
| Rank02 | ModelGO | Score59.0% | Percentile95.0% | Participants21 | EvidenceC | Evaluated |
| Rank03 | ModelBY | Score57.8% | Percentile90.0% | Participants21 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score53.5% | Percentile85.0% | Participants21 | EvidenceC | Evaluated |
| Rank05 | ModelMA | Score52.2% | Percentile80.0% | Participants21 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score51.3% | Percentile75.0% | Participants21 | EvidenceC | Evaluated |
| Rank07 | ModelTM | Score48.7% | Percentile70.0% | Participants21 | EvidenceC | Evaluated |
| Rank08 | ModelMA | Score48.7% | Percentile65.0% | Participants21 | EvidenceC | Evaluated |
| Rank09 | ModelMA | Score44.8% | Percentile60.0% | Participants21 | EvidenceC | Evaluated |
| Rank10 | ModelNV | Score44.6% | Percentile55.0% | Participants21 | EvidenceC | Evaluated |
| Rank11 | ModelME | Score43.6% | Percentile50.0% | Participants21 | EvidenceC | Evaluated |
| Rank12 | ModelNV | Score42.0% | Percentile45.0% | Participants21 | EvidenceC | Evaluated |
| Rank13 | ModelZA | Score41.7% | Percentile40.0% | Participants21 | EvidenceC | Evaluated |
| Rank14 | ModelMI | Score39.0% | Percentile35.0% | Participants21 | EvidenceC | Evaluated |
| Rank15 | ModelCO | Score38.2% | Percentile30.0% | Participants21 | EvidenceC | Evaluated |
| Rank16 | ModelCO | Score38.0% | Percentile25.0% | Participants21 | EvidenceC | Evaluated |
| Rank17 | ModelIN | Score38.0% | Percentile20.0% | Participants21 | EvidenceC | Evaluated |
| Rank18 | ModelZA | Score37.3% | Percentile15.0% | Participants21 | EvidenceC | Evaluated |
| Rank19 | ModelMI | Score36.0% | Percentile10.0% | Participants21 | EvidenceC | Evaluated |
| Rank20 | ModelNV | Score33.3% | Percentile5.0% | Participants21 | EvidenceC | Evaluated |
| Rank21 | ModelNV | Score32.6% | Percentile0.0% | Participants21 | EvidenceC | Evaluated |
SciCode Highlights
The leading models and scores on this benchmark.
SciCode Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for SciCode
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis scicode AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01BYSeed 2.1 ProByteDance
- Score
- 59.8%
Strengths
- Ranks #1 of 21 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures SciCode, not total model capability
- 02GOGemini 3.1 ProGoogle
- Score
- 59.0%
- Price
- $2.0 input / $12 output per 1M tokens
- Speed
- Up to 23 tok/s via Google
Strengths
- Ranks #2 of 21 compared models
- 95th percentile on this benchmark
- C evidence result
Considerations
- This result measures SciCode, not total model capability
- 03BYSeed 2.1 TurboByteDance
- Score
- 57.8%
Strengths
- Ranks #3 of 21 compared models
- 90th percentile on this benchmark
- C evidence result
Considerations
- This result measures SciCode, not total model capability
- 04ACQwen3.7 MaxAlibaba Cloud / Qwen Team
- Score
- 53.5%
- Price
- $2.5 input / $7.5 output per 1M tokens
- Speed
- Up to 5.8 tok/s via Together
Strengths
- Ranks #4 of 21 compared models
- 85th percentile on this benchmark
- C evidence result
Considerations
- This result measures SciCode, not total model capability
- 05MAKimi K2.6Moonshot AI
- Score
- 52.2%
- Price
- $0.95 input / $4.0 output per 1M tokens
- Speed
- Up to 285 tok/s via Fireworks
Strengths
- Ranks #5 of 21 compared models
- 80th percentile on this benchmark
- C evidence result
Considerations
- This result measures SciCode, not total model capability
Selection summary
Best AI Models for SciCode
Seed 2.1 Pro currently leads SciCode with 59.8%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is SciCode?
What SciCode measures and how its scores work.
SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- SciCode
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- scicode|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about SciCode.
Which model scores highest on SciCode?
Seed 2.1 Pro is currently ranked first with 59.8%.
What does SciCode measure?
SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
21 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.