Pyyan / Compare / Inkling 2 Small vs DeepSeek V4-Pro
Inkling 2 Small vs DeepSeek V4-Pro
2 of 5Open-Weight Models · verified 14 Nov 2026
TInkling 2 SmallThinking Machines Labcurrent
DeepSeek V4-ProDeepSeekcurrent | Specification | Inkling 2 Small | DeepSeek V4-Pro |
|---|
| Summary | A 594B multimodal model with 17B active, trained natively in 4-bit NVFP4, under Apache 2.0. | Best all-round open model of 2026. |
|---|
| Context | 1M | — |
|---|
| Input ($/Mtok) | Weights only | — |
|---|
| Output ($/Mtok) | Weights only | — |
|---|
| Licence | Apache 2.0 | MIT |
|---|
| Max output | Not published | — |
|---|
| Parameters | 594B total, 17B active | — |
|---|
| Released | 5 Nov 2026, weights 12 Nov | — |
|---|
| Model ID | thinkingmachines/Inkling-2-Small | — |
|---|
| Category | Open-Weight Models | Open-Weight Models |
|---|
| Official | Thinking Machines Lab ↗ | DeepSeek ↗ |
|---|
Highlighted rows are where these differ.
Inkling 2 Small
- Weights published on Hugging Face on 12 November 2026, a week after a hosted-only launch on Tinker on 5 November
- Takes text and images, writes text, with a context window of up to 1M tokens
- Pretrained, midtrained and post-trained in NVFP4. The main checkpoint needs at least 500GB of GPU memory, such as 2 B300s or 4 H200s
- An MXFP8 checkpoint is also published and needs at least 800GB. The lab says it scores only 1 to 2 points higher
- Local serving recipes for SGLang, vLLM, TokenSpeed and Unsloth. The benchmark scores on the card are the lab's own
Best for self-hosting a large text and image model on a multi-GPU node, or fine-tuning it on Tinker.
Full spec sheet →
DeepSeek V4-Pro
- Tops open leaderboards on agentic coding and reasoning
- MIT licence — no conditions
- Pioneered aggressive cache-hit pricing at ~$0.07/M
- DeepSeek routes all V4-Pro requests to V4.1 Flash from 14 September 2026
Best for everything open.
Full spec sheet →