Thinking Machines: Inkling
Inkling is a model from Thinking Machines that accepts text, image, and audio inputs, which makes it a candidate for workflows mixing documents, screenshots, and recorded speech. It offers a 524,288-token context window and can return up to 471,859 tokens in a single completion, so long inputs and long outputs both fit within one request. It supports tool calling and a reasoning mode; structured output support and weight availability are not specified in the available data. Pricing sits at $1.00 per million input tokens and $4.05 per million output tokens, so output-heavy jobs will dominate cost. Its blended benchmark score of 37.2 comes from a single independent benchmark, where it posted 41.2 on the AI Index, and one data point is thin evidence for ranking it against alternatives. Teams that need audio input alongside very long context should shortlist it, but they should run their own evaluations before committing.
Benchmark results
Independent, published benchmarks. Blended score 37.2 across 1 benchmark, last refreshed 2026-10-19. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 41.2 |
- Model ID
- thinkingmachines/inkling
- Vendor
- thinkingmachines
- Released
- July 2026
- Tokenizer
- Other
- Input Modalities
- text, image, audio
- Output Modalities
- text
- Max Output
- 471,859 tokens
- Tool Calling
- ✓ supported
- Structured Output
- not supported
- Reasoning Mode
- ✓ supported
- Vision
- ✓ accepts images
- Audio
- ✓ accepts audio
- Moderated
- no
What it costs in practice
Computed from the current $1.00/M input and $4.05/M output rates. Run your own numbers →
| Job | Tokens | Cost |
|---|---|---|
| Summarize a 50-page report | 30k in / 1.5k out | $0.04 |
| Classify 1,000 customer emails | 500k in / 50k out | $0.70 |
| A month of a busy support chatbot | 5M in / 2M out | $13.10 |
Price & spec history
Tracked daily by PicksByModel since 2026-07-18.
| Date | Input /M | Output /M | Context |
|---|---|---|---|
| 2026-10-07 | $1.00 | $4.05 | 524,288 |
| 2026-10-02 | $0.95 | $4.05 | 524,288 |
| 2026-09-26 | $1.00 | $4.05 | 524,288 |
| 2026-09-01 | $1.00 | $4.05 | 1,048,576 |
| 2026-08-25 | $0.95 | $4.05 | 1,048,576 |
| 2026-08-23 | $1.00 | $4.05 | 1,048,576 |
| 2026-08-08 | $0.95 | $4.05 | 1,048,576 |
| 2026-07-24 | $1.00 | $4.05 | 1,048,576 |
| 2026-07-22 | $1.00 | $4.05 | 524,288 |
| 2026-07-18 | $1.00 | $4.05 | 1,048,576 |
Category rankings
Where Thinking Machines: Inkling places across the 2 categories it ranks in. How we rank →
| # | Category | Score |
|---|---|---|
| #16 | Audio SummarizationVoice · of 25 ranked | 144 |
| #18 | TTS ReplacementVoice · of 25 ranked | 115 |
Similar models
Quick answers
- How much does Thinking Machines: Inkling cost?
- $1.00 per million input tokens and $4.05 per million output tokens.
- What is Thinking Machines: Inkling's context window?
- 524,288 tokens, roughly 786 pages of text in a single request.
- Does Thinking Machines: Inkling support tool calling?
- It supports tool calling, a reasoning mode.
- Can Thinking Machines: Inkling process images?
- Yes, it accepts image input.