Skip to content
Model Pareto

Inkling

Thinking Machines LabOpen weightsthinkingmachines.ai ↗
Compare with nearest rivalsCompare with…
Input
$1.00/M tok
Output
$4.05/M tok
Blended
$1.762/M tok · 3:1
Speed
184tok/s
Context
1Mtokens

Open weights; no first-party per-token API, price is the Artificial Analysis third-party 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
16× B200 180GB
FP8 weights
Throughput
~45,766 tok/s
many requests batched
Price
$0.27–2.40 /M tok
busy → light use
On your own machine
Too large
needs data-center GPUs
Assumptions (5)
  • FP8 weights (975B params, 41B active per token) + 40% KV-cache headroom ≈ 1365 GB
  • 16× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~45,766 output tok/s aggregate at batch 1541 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~352,646 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
~#27 of 31 (estimated)~52.8%
  1. #1DeepSeek V4.1 Flash74.2%
  2. #2Grok 4.772.6%
  3. #26Nex-N2.5-Pro55.8%
  4. ~#27InklingEstimated52.8%
  5. #27Qwen3.8 Max (0902)51.0%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#8 of 20 (estimated)~76.3%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #7gpt-oss-20b77.7%
  4. ~#8InklingEstimated76.3%
  5. #8Mistral Small 464.0%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~last of 15 (estimated)below measured range (<41.1)
  1. #1Claude Opus 5.566.0
  2. #2Claude Fable 5.162.2
  3. #14GPT-6 Luna41.1
  4. ~#15InklingEstimated<41.1
No published score — estimated from related benchmarks

Writing & chat

MMMLUEstimated
~#3 of 5 (estimated)~88.0%
  1. #1Gemini 3.1 Pro92.6%
  2. #2Gemma 4 31B88.4%
  3. ~#3InklingEstimated88.0%
  4. #3Claude Haiku 4.583.0%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Verified
#28 of 531430 Elo
  1. #1Claude Opus 5.51501
  2. #2Claude Fable 5.1Few votes1496
  3. #27Qwen3.5 397B A17B1434
  4. #28Inkling1430
  5. #29Qwen3.8 27B1429
lmarena.ai · 2026-09-25
LMArena Text (Elo)Verified
#28 of 531440 Elo
  1. #1Claude Opus 5.51509
  2. #2Claude Fable 5.11501
  3. #27MiniMax-M31441
  4. #28Inkling1440
  5. #29Grok 4.71439
lmarena.ai · 2026-09-24

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2InklingEstimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#5 of 8 (estimated)~83.2%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #4Gemini 3.1 Pro85.3%
  4. ~#5InklingEstimated83.2%
  5. #5Claude Haiku 4.579.6%
No published score — estimated from related benchmarks

Hard reasoning

Agents

τ²-bench (Telecom)Estimated
~#33 of 116 (estimated)~79.7%
  1. #1Step 3.7 Flash98.5%
  2. #2Gemini 3.1 Pro95.6%
  3. #32Nova 2.0 Omni80.4%
  4. ~#33InklingEstimated79.7%
  5. #33LongCat Flash Lite79.5%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#10 of 12 (estimated)~69.4%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. #9Nex-N2.5-Mini71.2%
  4. ~#10InklingEstimated69.4%
  5. #10Muse 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. ~#5InklingEstimated<47.5%
No published score — estimated from related benchmarks