Skip to content
Model Pareto

Flint

Recursive SuperintelligenceOpen weightshuggingface.co ↗
Compare with nearest rivalsCompare with…
Input
—
Output
—
Speed
~110tok/sest.
Context
262Ktokens

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.

Hardware
1× H200 141GB
FP8 weights
Throughput
~8,511 tok/s
many requests batched
Price
$0.044–0.37 /M tok
busy → light use
On your own machine
RTX 5090 32GB (int4, 300+ tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • FP8 weights (39.8B params, 2.5B active per token) + 25% KV-cache headroom ≈ 50 GB
  • 1× H200 141GB at $3.59–$7.91/GPU-hour on-demand (2026-09-24)
  • ~8,511 output tok/s aggregate at batch 256 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~187,016 tok/s
  • Low = 75% utilization at the low GPU price; high = 20% at the high price

Benchmark scores

~ italic, dashed = no published score yet, estimated from related benchmarks.

Coding

DeepSWE v1.1Estimated
~#40 of 43 (estimated)~31.0%
  1. #1Gemini 4 Argon78.8%
  2. #2DeepSeek V4.1 Flash 101777.0%
  3. #39Nex-N2.5-Mini36.1%
  4. ~#40FlintEstimated31.0%
  5. #40Hy328.0%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#40 of 99 (estimated)~3.0%
  1. #1Kimi K3.1 Pro67.2%
  2. #2Claude Sonnet 5.563.6%
  3. #39Gemini 3.1 Pro4.0%
  4. ~#40FlintEstimated3.0%
  5. #40GPT-5.4 mini2.0%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#15 of 28 (estimated)~65.7%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #14Maple-Preview75.1%
  4. ~#15FlintEstimated65.7%
  5. #15Mistral Small 464.0%
No published score — estimated from related benchmarks
SciCodeEstimated
~#64 of 118 (estimated)~44.7%
  1. #1Claude Opus 5.566.9%
  2. #2Claude Fable 5.163.1%
  3. #63Qwen3.5 397B A17B44.8%
  4. ~#64FlintEstimated44.7%
  5. #64Ling-3.0-flash-VL44.2%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~#12 of 24 (estimated)~47.7
  1. #1Claude Opus 5.561.7
  2. #2Claude Opus 559.7
  3. #11GLM-5.348.9
  4. ~#12FlintEstimated47.7
  5. #12DeepSeek V4.1 Flash47.3
No published score — estimated from related benchmarks