$0.06 per 1M input tokens and $0.18 per 1M output tokens (cheapest host: DeepInfra).
There is no free API; the cheapest paid price is $0.06 per 1M input tokens. The weights are open, so you can run it yourself.
#104 of 334 in Chat & Knowledge, #75 of 132 in Coding, #91 of 133 in Agents & Tool Use, #81 of 133 in Vision & Documents. Strongest at: Science.
On Humanity's Last Exam, Ling 3.0 Flash scores 23.7 vs 8.7 — higher.
Through the API of DeepInfra, Novita.
262K tokens, with up to 33K output tokens.
Released 2026-08-04. It has been superseded by Ling 3.1 Flash.
Yes, the weights are open (MIT licence).
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
No, it accepts text only.
Ling 3.0 Flash is an open-weights language model from inclusionAI, released Aug 2026, the successor to Ling 2.6 1T.
GPT OSS 120B (cheaper), GPT-5.6 Luna (better), Qwen3.8 Flash (open weights).