Flint 1.0 Benchmarks, Pricing & Rankings
Recursive · 2026-10-21 · Open · Current
Snapshot 2026-11-06 · updated weekly
Flint 1.0 ranks #71 of 71 on the AMR leaderboard.
Flint 1.0 benchmark scores
Flint 1.0 price, speed & context
Input price
Free
per 1M tokens
Output price
Free
per 1M tokens
Speed
—
tok/s
Context
262K
tokens
About Flint 1.0
Flint 1.0 is Recursive’s open-weights MoE for coding agents (about 2.5B active of 40B total), built by its Eureka Machine autoresearch pipeline. Listed on OpenRouter October 21 as a free endpoint with a 256K context. Data source: OpenRouter listing only — no leaderboard, benchmark, or arena data yet, so AMR score and throughput stay unranked (0) pending first evaluations.
Models around Flint 1.0 in the rankings
Flint 1.0 FAQ
Where does Flint 1.0 rank?
Flint 1.0 ranks #71 of 71 tracked AI models on the AMR aggregate as of October 2026, with an aggregate score of 0.0. Its strongest dimension is AMR at 0.0.
How much does Flint 1.0 cost?
Flint 1.0 costs Free per million input tokens and Free per million output tokens on the standard API. At a typical 1:3 input/output agent workload, that blends to roughly $0.00 per million tokens.
What is the Flint 1.0 context window?
Flint 1.0 supports a 262K token context window and generates at a median 0 output tokens per second. That context fits roughly 175 standard documents in a single prompt.
Is Flint 1.0 open source?
Yes — Flint 1.0 ships open weights that you can download and self-host or serve through any compatible inference provider.
What is better than Flint 1.0?
Mercury 3 ranks one place higher at #70 with an AMR score of 0.0. Flint 1.0 still wins 0 of the tracked dimensions head-to-head — check the comparison table to see exactly which.