Seed2.1
Seed2.1 is capable in multimodal tasks and behind the leaders in agentic tasks and coding. Too few results yet to rate reasoning, safety, long context, math, multilingual tasks, instruction following, or factuality.
- Price per million tokens
No current price is tracked for this model. See the rate card
- Evidence
50results on50benchmarks
- 50 vendor-reported
From 3 sources · latest Sep 28, 2026 · How verification works
Capability profile
Bars show the model's median result as a share of the leading model's, per capability.
Too few results to rate: ReasoningSafetyLong ContextMathMultilingualInstruction FollowingFactuality
Seed2.1 benchmark results
50 results on 50 benchmarks, grouped by capability. Every score links to its source; a bar is the result as a share of the capability leader's.
Multimodal
Capable3 of 6 ranked benchmarks measuredFull multimodal ranking- 90.50100%
- Video-MME89.00100%VideoMME
- 80.1091%
2 more multimodal results
- 79.40—
- OCRBench v262.80—OCRBenchv2
Agentic
Limited2 of 7 ranked benchmarks measuredFull agentic ranking- GDPval (win rate)82.7094%GDPval
- MCP Atlas80.3091%MCP-Atlas
1 more agentic result
- AA ApexAgents29.20—APEX Agents
Coding
Limited2 of 10 ranked benchmarks measuredFull coding ranking- SciCode57.8086%SciCode
- 67.6074%
Reasoning
Not enough data0 of 6 ranked benchmarks measuredFull reasoning ranking- ARC-AGI-261.30—ArcAGI2
More results
- Code Arena: Frontend1539.00—
- BrowseComp (with Search)84.90—
- 54.60—
- 42.00—
- 88.00—
- MathArena Apex (Pass@1)35.40—MathArena Apex
33 more results
- BabyVision62.90—BabyVision
- 70.90—
- CyberGym67.00—CyberGym
- 23.00—
- 83.10—
- 68.10—
- EmbSpatial-Bench82.50—EmbSpatialBench
- 71.30—
- Finance Agent v1.156.00—
- HLE-Verified (no tool)42.40—
- 80.60—
- 76.80—
- MathVerse (Vision-Only)89.20—
- 65.90—
- 76.90—
- MMSIBench (circular)31.40—
- 74.80—
- Office QA Pro [Multimodal]71.10—
- OfficeQA Pro62.80—OfficeQA-Pro
- 69.70—
- 86.30—
- 71.10—
- SuperGPQA67.40—SuperGPQA
- SWE-Pro Bench57.00—
- 56.80—
- Toolathlon49.10—Toolathlon
- 77.20—
- 67.60—
- 71.40—
- 52.90—
- 48.60—
- ZeroBench (main)11.00—
- ZeroBench (sub)49.10—
About this record
Verification: 50 scores · 0 independently verified · 50 vendor-reported. How these tiers are assigned
- Tracked since
- Sep 28, 2026
- Newest source mention
- Sep 28, 2026