thinkingmachines
Thinking Machines: Inkling 2 Small (free)
thinkingmachines/inkling-2-small:free
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Inkling 2 Small is an open-weight multimodal mixture-of-experts reasoning model from Thinking Machines Lab, documented by its model card as a 48-layer transformer with 594B total and 17B active parameters that accepts text and image input and emits text. The card describes a context window of up to 1M tokens and English-first multilingual coverage, while the OpenRouter catalog lists this free variant with mandatory reasoning and tool support. Benchmark figures are vendor-published; independent verification was not available in this source pack.[1][2]
Strengths
- Vendor-documented sparse MoE design activates only 17B of 594B parameters per token, and both sources report a context window of up to 1M tokens.[1][2]
- Natively multimodal per the model card, with images encoded by a vision transformer into a shared decoder pipeline; OpenRouter lists tool calling and adjustable reasoning effort among supported parameters.[1][2]
- Open weights under an Apache-2.0 license, with documented local inference via SGLang, vLLM, TokenSpeed or Unsloth and post-training options such as Miles and SkyRL.[2]
Limitations
- Self-hosting demands a multi-GPU cluster: the card specifies at least 500GB of aggregated VRAM for the recommended NVFP4 checkpoint and 800GB for the MXFP8 variant.[2]
- Per the OpenRouter catalog, reasoning is mandatory on this endpoint and cannot be disabled, although several effort levels are configurable.[1]
- Evidence gaps: knowledge cutoff is not stated, benchmark results are vendor-published rather than independently verified, and real-world quality should be validated on your own workloads.[1][2]
Where to use it
- A candidate for agentic and tool-use systems, coding assistants, chatbots and retrieval-augmented generation, which the model card lists as intended applications; validate on your own tasks.[2]
- Worth evaluating for long-document or image-plus-text conversational workloads, given the reported 1M-token context and native image input.[1][2]
- Potentially attractive for zero-cost experimentation through this OpenRouter free variant, or for fine-tuning research enabled by the open-weight release.[1][2]
Sources
- [1] OpenRouter catalog · Thinking Machines: Inkling 2 Small (free)Sources checked: Nov 13, 2026
- [2] Model documentation · Thinking Machines: Inkling 2 Small (free)Sources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Inkling 2 Small is an open-weight multimodal mixture-of-experts reasoning model from Thinking Machines Lab, with 17B active parameters out of 594B total. It is the first model in the second...
Specifications
- Model ID
- thinkingmachines/inkling-2-small:free
- Canonical ID
- thinkingmachines/inkling-2-small-20261107
- Creator
- thinkingmachines
- Added to OpenRouter
- 09.11.2026
- Context length
- 1,048,576
- Modality
- text+image->text
- Input modalities
- text, image
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0 / out $0
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | BB — next 30% of the cohort | reasoning: mandatory · default high · efforts: max, high, medium, low, minimal · parameters: include_reasoning, reasoning, reasoning_effort |
| Coding | No measurement | — |
| Agents | No measurement | parameters: tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: mandatory · default high · efforts: max, high, medium, low, minimal · parameters: include_reasoning, reasoning, reasoning_effort |
| Document analysis | No measurement | — |
| Vision | No measurement | input: image |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Intelligence index | 30.6 | BB — next 30% of the cohort | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Thinking Machinesnvfp4 | 1,048,576 | $0 | $0 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.