Vedant
554 posts
This week is crazy for ai model releases.
Everybody is releasing a model even meta did lol
Everyone except google
0xMarioNawfal@RoundtableSpace
Singapore just dropped a free model that's competing with frontier numbers. > Agnes 2.5 Pro: 82.7 on SWE-bench Verified, 78.7 on Multilingual > Agnes 2.5 Flash improved on 2.0 Flash across every single benchmark > Biggest gains on SWE Atlas - beating GLM 5.2 and DeepSeek V4 Pro on several cuts - Free API available today - Built by a team focused on proper multimodal from the ground up The frontier is no longer just the US and China.
English
Gemini 3.1 Pro High in Antigravity 🥲
> Simulation of Long March 10B booster approaching the net-capture tower with three js
Harshith@HarshithLucky3
GPT 5.6 sol high vs Claude Fable 5 high > Simulation of Long March 10B booster approaching the net-capture tower with three js both nailed it 🔥🔥
English
This is why competition is great
Claude@claudeai
We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.
English
muse spark 1.1 outperforms opus, grok 4.5, and gemini on a new challenging finite model theory / theoretical cs eval
Serafim Batzoglou@s_batzoglou
I benchmarked the new models (Sol, Terra, Luna, Fable 5, Meta Muse Spark 1.1, Grok 4.5) on an induction reasoning task. This is an updated table from yesterday, and my benchmark is described as spotlight in ICML '26. How to read the table: The models are given several small graphs (6-10 graphs, each with 8-10 elements), in which some nodes are designated. They must return a first-order formula that describes properties of the designated nodes simultaneously for all graphs. The benchmark has 64 such problems (a subset of the bigger benchmark). Correct: how many formulas correctly describe the designated nodes. Holdout correct: a number of additional graphs, with the same underlying property, are held out. This is to test the ability of the model's correct formulas to generalize. For instance, if a formula is a big case split rather than the underlying simple graph property, it won't generalize. Formula complexity (in AST): mean and median size of the correct model's formulas. A few observations and experiment notes: - All new models are remarkably good. Even Luna and Muse Spark outperforms GPT 5.4. - Some models are better in returning simple formulas that generalize well. - Fable 5 was extremely hard to get results from. I first ran it with higher thinking effort levels, in which case it charged for max tokens per problem and returned all "" responses. The only settings in which it returns non-"" responses are medium and low. My method for running it was to call it on medium, then call it again on medium, then on low. - GPT 5.5 is absent because it had the same non-return behavior as Fable 5 above, but even worse. So bad that I couldn't get almost any results and would have wasted too much $$ on empty responses. Happy to see that the new OpenAI models are "cured" of this GPT 5.5 issue. - I had observed that Grok 4 >> Grok 4.1 Fast > Grok 4.20 >> Grok 4.3. Finally, this downward trend of XAI has been reversed and Grok 4.5 is now a contender. - Meta Muse Spark 1.1 is a big surprise. It beat GPT 5.4 on this task! - The paper is out and the benchmark is public: icml.cc/virtual/2026/e…
English
Pricing is definitely the best part of muse spart 1.1
Alexandr Wang@alexandr_wang
muse spark 1.1 outperforms opus, grok 4.5, and gemini on a new challenging finite model theory / theoretical cs eval
English
Vedant retweetledi
Grok 4.5 and Meta Muse Spark 1.1 are both free to try right now 😳
no credit card
no subscriptions to get started
just pick a model and start building
what you get for $0:
- Grok 4.5 with strong coding and agent performance
- Meta Muse Spark 1.1 with advanced tool use and computer tasks
- free access in Cursor and Meta AI
- $20 free API credits from Meta
- OpenAI-compatible APIs for your existing tools
why this is huge:
> no paying for Claude or GPT just to test frontier models
> works with the tools you already use
> compare both models before spending on API credits
getting started takes less than 5 minutes:
Grok 4.5
1. open cursor.com
2. select Grok 4.5 from the model list
3. or use build.grok.com
4. start coding for free
Meta Muse Spark 1.1
1. go to meta.ai
2. enable Thinking Mode
3. visit developer.meta.com/ai/products/me…
4. claim your $20 free API credits
5. connect it to your favorite OpenAI-compatible client
works with:
> Cursor
> Cline
> OpenCode
> Replit
> any OpenAI-compatible tool
pro tip:
use Grok 4.5 for coding and debugging, then compare the same tasks with Muse Spark 1.1 for tool use and agent workflows
the best part?
one model runs directly inside Cursor, while the other is free in Meta AI and includes API credits for developers.
two frontier AI models, free access, and $20 in API credits without paying for Claude or GPT
bookmark this before the launch offers change 👀
English
Which AI do you code with the most?
- Claude
- ChatGPT
- Grok
- Gemini
English
I still can’t figure out why Gemini struggles to compete with Claude and GPT.
- Owns Chrome
- Backed by Android
- Stores most search results
- Holds ~95% search history
- Google has the biggest user data
- Even incognito data isn’t fully private
So what’s the problem?
English
@ritu_twts Grok 4.5 for planning
Composer 2.5 for implementation
This is the best combination you can use right now, the price to performance ratio is amazing
English
