$ diff astra inkling
Astra vs Inkling
Values come from Astra and Inkling, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
OpenAIAstra
- context
- 1.1M
- weights
- closed
- $/M in
- $10
- $/M out
- $50
- context
- 1M
- weights
- open
- $/M in
- —
- $/M out
- —
What actually differs
- Context window
- Astra takes 1.1M against 1M — modestly more room in a single request.
- Weights
- Inkling publishes weights (Apache 2.0 (open-weight)); Astra is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- Astra shipped 50 days after Inkling (2026-09-03 vs 2026-07-15).
Full spec
| Attribute | Astra | Inkling |
|---|---|---|
| Developer | OpenAI | Thinking Machines |
| Released | 2026-09-03 | 2026-07-15 |
| Context window | 1,050,000 tokens (128,000 max output) | 1M |
| Pricing | $10/M input · $50/M output | not recorded |
| License | proprietary | Apache 2.0 (open-weight) |
| Availability | ChatGPT Plus, Pro, Business, Enterprise · OpenAI API · AWS — staged, see below | Hugging Face (BF16 and NVFP4), Tinker (fine-tuning), SGLang, vLLM, llama.cpp, Unsloth |