Skip to content
Model Pareto

Inkling Small

Thinking Machines LabOpen weightsartificialanalysis.ai ↗
Input
$0.30/M tok
Output
$1.20/M tok
Blended
$0.525/M tok · 3:1
Speed
206tok/s
Context
1Mtokens
Cost per task
~$0.51/taskest.

Long-tail import from Artificial Analysis (2026-09-24); AA variant: Inkling Small. Price = AA median host price.

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.

Hardware
4× B200 180GB
FP8 weights
Throughput
~22,891 tok/s
many requests batched
Price
$0.12–1.03 /M tok
busy → light use
On your own machine
Mac Studio M5 Ultra 512GB (int4, ~140 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • FP8 weights (276B params, 12B active per token) + 40% KV-cache headroom ≈ 386 GB
  • 4× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~22,891 output tok/s aggregate at batch 655 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~354,375 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

~#9 of 21 (estimated)~76.9%
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. #7Mistral Medium 3.577.6%
  4. ~#9Inkling SmallEstimated76.9%
  5. #9Ripple-65B Preview76.5%
No published score — estimated from related benchmarks
~#17 of 24 (estimated)~56.7%
  1. #1Claude Fable 5.293.0%
  2. #2Claude Opus 5.589.9%
  3. #16Hy357.9%
  4. ~#17Inkling SmallEstimated56.7%
  5. #17Inkling54.3%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#37 of 46 (estimated)~52.6%
  1. #1Gemini 4 Argon78.8%
  2. #2DeepSeek V4.1 Flash 101777.0%
  3. #36Nex-N2.5-Pro55.8%
  4. ~#37Inkling SmallEstimated52.6%
  5. #37Qwen3.8 Max (0902)51.0%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#15 of 28 (estimated)~73.4%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #14Maple-Preview75.1%
  4. ~#15Inkling SmallEstimated73.4%
  5. #15Mistral Small 464.0%
No published score — estimated from related benchmarks
~#23 of 26 (estimated)~33.9
  1. #1Claude Opus 5.561.7
  2. #2Claude Opus 559.7
  3. #22Claude Haiku 5.537.3
  4. ~#23Inkling SmallEstimated33.9
  5. #23Gemini 3.8 Flash33.3
No published score — estimated from related benchmarks

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2Inkling SmallEstimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
~#10 of 12 (estimated)~77.9%
  1. #1Gemini 3.8 Flash86.2%
  2. #2Kimi K384.8%
  3. #9Inkling78.1%
  4. ~#10Inkling SmallEstimated77.9%
  5. #10Claude Sonnet 577.0%
No published score — estimated from related benchmarks
~#5 of 8 (estimated)~84.1%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #4Gemini 3.1 Pro85.3%
  4. ~#5Inkling SmallEstimated84.1%
  5. #5Claude Haiku 4.579.6%
No published score — estimated from related benchmarks
GDP.pdf (AA)Estimated
~#29 of 48 (estimated)~12.9%
  1. #1Muse Wonder Full32.6%
  2. #2GPT-6 Astra31.0%
  3. #28Claude Sonnet 513.2%
  4. ~#29Inkling SmallEstimated12.9%
  5. #29DeepSeek V4.1 Flash12.8%
No published score — estimated from related benchmarks
ChartographyEstimated
~#17 of 25 (estimated)~18.8%
  1. #1Claude Fable 5.273.4%
  2. #2GPT-6 Astra71.0%
  3. #16Grok 4.620.5%
  4. ~#17Inkling SmallEstimated18.8%
  5. #17Muse Glimmer17.2%
No published score — estimated from related benchmarks

Hard reasoning

Agents

~#11 of 15 (estimated)~67.6%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. #10Nex-N2.5-Mini71.2%
  4. ~#11Inkling SmallEstimated67.6%
  5. #11Muse Glimmer65.9%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. #4GPT-5.547.5%
  4. ~#5Inkling SmallEstimated<47.5%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#19 of 22 (estimated)~67.7%
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #18Nemotron 3.5 Bolt 74B-A9B70.0%
  4. ~#19Inkling SmallEstimated67.7%
  5. #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
~#29 of 92 (estimated)~1380 Elo
  1. #1Claude Opus 5.51866
  2. #2Muse Wonder Full1857
  3. #28Qwen3.8 27B1419
  4. ~#29Inkling SmallEstimated1380
  5. #29K2 Horizon 375B A23B1361
No published score — estimated from related benchmarks
~#38 of 85 (estimated)~23.9%
  1. #1Claude Fable 5.276.4%
  2. #2Muse Wonder Full71.3%
  3. #37Kimi K2.7 Code24.5%
  4. ~#38Inkling SmallEstimated23.9%
  5. #38Quasar 438B23.9%
No published score — estimated from related benchmarks