thinkingmachines/inklingInkling is Thinking Machines’s open-weights AI model, released Jul 15, 2026. It costs $0.95/M input and $4.05/M output tokens, and handles a 1M-token context window.
Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total.
Scores on standardized evaluations. Higher is better — and rank shows where Inkling lands among all models tracked on Model Beat.
Abstract visual reasoning
Harder abstract reasoning
Graduate-level scientific reasoning
Frontier of human expert knowledge
Common-sense trick questions
Factual accuracy & hallucination
Novel ML problem-solving
Scientific research coding
Human-rated web development
Olympiad-qualifier math
Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total.
Inkling is an AI model developed by Thinking Machines, released Jul 15, 2026.
Inkling is served from $0.95 per million input tokens and $4.05 per million output tokens (cheapest credible provider on OpenRouter: DeepInfra).
Inkling supports a context window of up to 1M tokens.
On standardized evaluations tracked by Epoch AI, Inkling scores 88.3% on GPQA Diamond, 31.9% on Humanity's Last Exam, 47.0% on SciCode.
Inkling is released with open weights (Open weights (unrestricted)), so the model can be downloaded and self-hosted.
Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Sep 10, 2026.