Inkling
975B-parameter Apache 2.0 open model built as a fine-tuning foundation.
About
The first model from Mira Murati's Thinking Machines Lab: a 975B-parameter MoE (41B active) pretrained on 45 trillion tokens of text, image, audio, and video, with a 1M context and Apache 2.0 weights. It reads text, images, and audio but outputs text only, and exposes a thinking-effort dial (0.2–0.99) that trades reasoning depth for token spend.
The lab is unusually candid that Inkling is 'not the strongest overall model available' — its pitch is being the best open base to fine-tune, natively on the lab's Tinker platform.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
- swe-bench-verified: 77.6 (lab-reported)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Thinking Machines Lab's first model, July 15, 2026; 975B MoE / 41B active, 1M context; Pretrained on 45T tokens across text, image, audio, video; outputs text only; Lab's own positioning is honest: not the strongest model, built as a fine-tuning base for Tinker; SWE-Bench Verified 77.6% is lab-reported. License: Apache 2.0 — full commercial use permitted..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · Thinking Machines Lab's first model, July 15, 2026; 975B MoE / 41B active, 1M context
- · Pretrained on 45T tokens across text, image, audio, video; outputs text only
- · Lab's own positioning is honest: not the strongest model, built as a fine-tuning base for Tinker
- · SWE-Bench Verified 77.6% is lab-reported