Inkling
Inkling is a large language model made by Thinking Machines, added to the catalog on 17 Jul 2026. It has a context window of 524,288 tokens and costs $1.00 per 1M input tokens and $4.05 per 1M output tokens. It accepts text, images and audio and produces text. Typical uses: reasoning over long text inputs, coding assistance and analyzing images with text output.
Price checked ยท source: OpenRouter
What Inkling is
Inkling is an open-weight multimodal model from Thinking Machines Lab. It accepts text, image, and audio inputs and produces text. The catalog describes it as a general-purpose model for reasoning, coding, and agentic and tool-use systems. It supports reasoning and tools, making it relevant to applications that combine model responses with tool-supported steps. The provided facts do not identify particular tools, input formats, or task-specific configurations, so those details depend on the system in which Inkling is used. Its listed output modality is text; image and audio are listed as input modalities.
Inkling has a context window of 524,288 tokens and a maximum output length of 471,859 tokens. The context capacity can accommodate prompts with substantial amounts of material, while the output limit allows for long text responses. The facts also identify it as a mixture-of-experts model with 41 billion active parameters out of 975 billion total. These details may help teams assess the model for deployments that need multimodal input, reasoning, coding assistance, or tool use over extensive context. The catalog does not specify supported languages, image or audio formats, or implementation requirements. Potential tasks include coding assistance, working with long documents, using images or audio as inputs to text-response workflows, and building tool-supported applications. Its text output suits applications where responses are presented as written content.
Key features
- Accepts text, image, and audio inputs and produces text
- Supports reasoning and tools
- Has a context window of 524,288 tokens
- Supports outputs of up to 471,859 tokens
- Described for general-purpose reasoning, coding, and tool-use systems
Typical uses
- Reasoning over long text inputs
- Coding assistance
- Analyzing images with text output
- Working with audio inputs
- Tool-supported workflows
Inkling pricing
US dollars, list price through OpenRouter.
| Item | Standard |
|---|---|
| Input, per 1M tokens | $1.00 |
| Output, per 1M tokens | $4.05 |
| Cached input read, per 1M tokens | $0.17 |
No price change since it was added on 17 Jul 2026.
| What a task costs | Cost |
|---|---|
| Answer a short question500 in / 300 out | $0.0017 |
| Summarise a 50-page PDF25,000 in / 800 out | $0.028 |
| Review a 2,000-line source file30,000 in / 2,000 out | $0.038 |
| Read a 300-page contract set150,000 in / 1,500 out | $0.156 |
Inkling specifications
| Model id | thinkingmachines/inkling |
|---|---|
| Version | inkling-20260715 |
| Context window | 524,288 tokensAbout 393,216 English words |
| Max output | 471,859 tokens |
| Knowledge cutoff | Not published |
| Added to catalog | 17 Jul 2026 |
| Input types | text, image, audio |
|---|---|
| Output types | text |
| Tool calling | Yes |
| Structured outputs | No |
| Reasoning effort control | Yes |
| Content moderation | Not applied by provider |
Sample outputs
The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.
No output has been recorded for Inkling yet.
Compare Inkling
Side by side with other models: prices, context window, what each accepts and supports, and the same four tasks priced on both.