DeepSeek Harness Benchmarks, Pricing & Rankings
DeepSeek · 2026-08-13 · Open · Current
Snapshot 2026-10-14 · updated weekly
DeepSeek Harness ranks #14 of 42 on the AMR leaderboard.
DeepSeek Harness benchmark scores
| GPQA Diamond | 86.5% | |
| MMLU-Pro | 84% | |
| SWE-Bench Verified | 71.2% | |
| AIME 2025 | 89.4% | |
| LiveCodeBench v6 | 77.6% | |
| Arena Elo | 1418 |
DeepSeek Harness price, speed & context
Input price
$0.35
per 1M tokens
Output price
$1.40
per 1M tokens
Speed
105
tok/s
Context
256K
tokens
About DeepSeek Harness
DeepSeek Harness is the newest open-weights flagship from DeepSeek, built for agentic tool use, and it undercuts every closed top-15 model on price by an order of magnitude.
Models around DeepSeek Harness in the rankings
More DeepSeek models
DeepSeek Harness FAQ
Where does DeepSeek Harness rank?
DeepSeek Harness ranks #14 of 42 tracked AI models on the AMR aggregate as of October 2026, with an aggregate score of 88.8. Its strongest dimension is AIME 2025 at 89.4%.
How much does DeepSeek Harness cost?
DeepSeek Harness costs $0.35 per million input tokens and $1.40 per million output tokens on the standard API. At a typical 1:3 input/output agent workload, that blends to roughly $1.14 per million tokens.
What is the DeepSeek Harness context window?
DeepSeek Harness supports a 256K token context window and generates at a median 105 output tokens per second. That context fits roughly 171 standard documents in a single prompt.
Is DeepSeek Harness open source?
Yes — DeepSeek Harness ships open weights that you can download and self-host or serve through any compatible inference provider.
What is better than DeepSeek Harness?
Gemini 4.0 ranks one place higher at #13 with an AMR score of 89.0. DeepSeek Harness still wins 3 of the tracked dimensions head-to-head — check the comparison table to see exactly which.