T
Thinking Machines Lab
San Francisco, CA-based AI company founded 2025. Private. Operates in AI/ML infrastructure. Serves researchers and developers with Tinker, an API for distributed LLM fine-tuning using LoRA, SFT, RL, and distillation.
Plans tracked
31
1 add-on
Pricing changes
1
Latest Wk 30 '26
Pricing model
Usage-based
No free tier
Pricing Change History
1 changeResearch Thinking Machines Lab's pricing with the Pulse Agent
Ask anything about Thinking Machines Lab's plans, packaging shifts, and how it compares across the market — answered live from verified data.
Ask the agentCurrent Plans
AnnualMonthly
Nemotron-3-Super-120B-A12B-BF16 (64K)
Plan
$1.44 per million sample tokens + $0.57 per million prefill tokens
Nemotron-3-Super-120B-A12B-BF16 (256K)
Plan
$1.92 per million sample tokens + $0.76 per million prefill tokens
Nemotron-3-Nano-30B-A3B-BF16 (64K)
Plan
$0.49 per million sample tokens + $0.2 per million prefill tokens
GLM-5.3 (256K)
Plan
$12.15 per million sample tokens + $0.97 per million cached prefill tokens
Qwen3.8-27B (64K)
Plan
$5.59 per million sample tokens + $0.37 per million cached prefill tokens
Qwen3.6-27B (64K)
Plan
$5.59 per million sample tokens + $0.37 per million cached prefill tokens
Qwen3.5-35B-A3B-Base (64K)
Plan
$1.33 per million sample tokens + $0.11 per million cached prefill tokens
Qwen3.5-4B (64K)
Plan
$1 per million sample tokens + $0.33 per million prefill tokens
GPT-OSS-120B (128K)
Plan
$0.78 per million prefill tokens + $2.33 per million train tokens
Inkling-Small Serverless Inference (256K)
Plan
$0.3 per million prefill tokens + $0.06 per million cached prefill tokens
Kimi K2.6 (32K)
Plan
$2.21 per million prefill tokens + $4.84 per million train tokens
Qwen3.8-27B (256K)
Plan
$2.48 per million prefill tokens + $7.46 per million train tokens
Qwen3.5-397B-A17B (64K)
Plan
$3 per million prefill tokens + $6.6 per million train tokens
Kimi K2.6 (128K)
Plan
$1.03 per million cached prefill tokens + $5.15 per million prefill tokens
Qwen3.6-35B-A3B (64K)
Plan
$0.11 per million cached prefill tokens + $0.54 per million prefill tokens
Qwen3.5-397B-A17B (256K)
Plan
$0.8 per million cached prefill tokens + $4 per million prefill tokens
Qwen3.5-9B-Base (64K)
Plan
$0.66 per million prefill tokens + $1.46 per million train tokens
GPT-OSS-120B (32K)
Plan
$0.84 per million sample tokens + $0.33 per million prefill tokens
DeepSeek-V3.1 (32K)
Plan
$4.21 per million sample tokens + $1.7 per million prefill tokens
Inkling Serverless Inference (256K)
Plan
$4.05 per million sample tokens + $1 per million prefill tokens
Qwen3.5-9B (64K)
Plan
$2 per million sample tokens + $0.66 per million prefill tokens
Qwen3-8B (32K)
Plan
$0.04 per million cached prefill tokens + $0.6 per million sample tokens
GPT-OSS-20B (32K)
Plan
$0.04 per million cached prefill tokens + $0.45 per million sample tokens
Inkling (64K)
Plan
$4.68 per million sample tokens + $1.87 per million prefill tokens
Inkling (256K)
Plan
$9.36 per million sample tokens + $3.74 per million prefill tokens
Inkling-Small (64K)
Plan
$1.44 per million sample tokens + $0.58 per million prefill tokens
Inkling-Small (256K)
Plan
$2.89 per million sample tokens + $1.16 per million prefill tokens
Nemotron-3.5-Lightning-30B-A3B-BF16 (64K)
Plan
$0.49 per million sample tokens + $0.2 per million prefill tokens
Nemotron-3.5-Lightning-30B-A3B-BF16 (256K)
Plan
$0.66 per million sample tokens + $0.26 per million prefill tokens
Nemotron-3-Ultra-550B-A55B-BF16 (64K)
Plan
$6.22 per million sample tokens + $2.49 per million prefill tokens
Nemotron-3-Ultra-550B-A55B-BF16 (256K)
Plan
$8.3 per million sample tokens + $3.32 per million prefill tokens
Storage
Add-on
$0.1 per GB per month
Company Facts
CategoryAI & Machine Learning
Free tierNo
Employees11-50
Tracked since2026-06-24
Pricing last seen2026-09-09
Websitethinkingmachines.ai
