Frontier, p. 4
Frontier
76 stories · page 4 of 9DeepSeek released the weights of its vision model under MIT and this time showed the numbers
On 31 August DeepSeek uploaded DeepSeek-V4-Flash-Vision-Exp to Hugging Face under the MIT licence. Ten days earlier the model was API-only, without a single image benchmark. Now the table is out, comparing the new model with their older one and with Opus 4.8.
Read →The new Qwen runs 6 billion parameters out of 125 - and its licence is no longer Apache
Alibaba announced Qwen3.8-Flash on 27 August as a preview of the architecture behind Qwen4. The weights really are free to download, but they are named Qwen3.8-Flash-Next and come under a community licence that requires a separate agreement if you build a coding or office assistant.
Read →Jalapeño's first numbers: OpenAI claims its chip gives more and faster at once
OpenAI published the first measurements of its own inference chip. They claim higher throughput and shorter waiting at the same time, which until now was treated as a trade. The numbers are theirs and were measured by them.
Read →Mistral and HUMAIN are building sovereign AI for Saudi Arabia
Mistral and Saudi Arabia's HUMAIN announced joint work on sovereign artificial intelligence for Saudi Arabia and the Middle East: infrastructure, models, deployment, in the range of hundreds of millions of euros. The computing stays inside HUMAIN's data centres.
Read →DeepSeek shipped vision over the API and counts an image as up to 384 tokens
An experimental variant of their fast model now understands images too, not only text. You pay for it at the text model's rate. For the promised jump on the multimodal agentic tests there is not a single published number.
Read →Grok 4.6 entered Google's enterprise platform
Since 21 August xAI's model sits inside Google's Gemini Enterprise Agent Platform, available through Model Garden. A 500,000 token context, adjustable reasoning level, and customers who would otherwise never open an account with xAI.
Read →DeepMind is teaching AI agents inside the world of EVE Online
Google DeepMind announced a research partnership with Fenris Creations, the studio behind EVE Online. Three game environments for long memory, planning, and agents that adapt. The work starts offline, and only mature capabilities reach live players.
Read →Grok Bot gives agents their own computer and lets them work around the clock
SpaceXAI released, in early beta, 'AI teammates' that sign into your apps, finish the job and only come back for approval. For SuperGrok Heavy and Cursor Ultra/Teams Premium subscribers.
Read →Anthropic put agents against each other and they got along too well
In the pricing test the agents settle the moment you give them a private channel. In another, 18 out of 30 create a branch with exactly the same name.
Read →