Sabitlenmiş Tweet
AI/ML API
1.6K posts
AI/ML API
@aimlapi
AI/ML API is an AI platform for developers and SaaS entrepreneurs looking to integrate cutting-edge AI capabilities into their products.
Katılım Şubat 2024
261 Takip Edilen1.7K Takipçiler
@eliasdevs @AlexFromAtomic @howlshot @astracollab Send email you used while registering to @schwarzerrttr DM
English
English
I made a research about new Alibaba frontier model and I am SHOCKED. Qwen 3.7 - Max is the greatest model now. Switch all my agents over to it..
atomic.chat@atomic_chat_hq
Qwen 3.7-max beats Opus 4.7 and GPT-5.5 We tested three frontier models on a real agentic task: write a Tetris bot that plays the game and trains itself. Each model could read its own code, run benchmarks, and rewrite itself across 10 iterations. Then we compared the final bots head to head. Qwen 3.7-Max: training cost $1.32, bot improvement +56% Claude Opus 4.7: training cost $12.15, bot improvement +28% GPT-5.5: training cost $2.85, bot improvement +7% Qwen won on every dimension - biggest jump, 9× cheaper than Claude, 2× cheaper than GPT. Long agentic loops is where Qwen Max actually delivers.
English
English
@aimlapi @AlexFromAtomic @howlshot Should I switch @astracollab to you guys then 🥹. We currently use open outer
English
AI/ML API retweetledi
@ECLresearch @ai_for_success We have not compared it yet with Sonnet - but I would suggest that it may well outperform it
English
@ai_for_success The 56.6 A² index is impressive, but the real value is in the reduced hallucination rate for agentic workflows — that’s where enterprise trust breaks or builds. How does it perform on multi-step tool-calling tasks vs. Claude Sonnet 4?
English
Qwen just launched Qwen 3.7 Max, and it’s now available on AI/ML API.
Qwen 3.7 Max scored 56.6 on the Artificial Analysis Intelligence Index and shows much stronger agentic performance.
It’s especially great for coding and tends to hallucinate less compared to many other models.
AI/ML is also giving away free promo codes to try it out. Check below 👇
AI/ML API@aimlapi
Qwen3.7-Max on AI/ML API - built for the agent era GPQA Diamond (92.4), HMMT (97.1), Apex (44.5) Sustains 35+ hours of autonomous execution Works with Claude Code, Qwen Code & more Comment Qwen to get Free promo code
English
@0xWeq @rohanpaul_ai We don't think so - we want everyone to experience great models, and that's it 😎
English
Qwen 3.7 Max is super close to the frontier models for coding and agentic abilities.
And and it’s now available on AI/ML API.
Agent reliability the center of the story and also on Artificial Analysis it's sitting at 5th, pretty much on par with GPT 5.4 (xhigh) and a notch above the just released Gemini 3.5 Flash.
AI/ML API is also giving away free codes for users who want to try it. see the quoted tweet.
AI/ML API@aimlapi
Qwen3.7-Max on AI/ML API - built for the agent era GPQA Diamond (92.4), HMMT (97.1), Apex (44.5) Sustains 35+ hours of autonomous execution Works with Claude Code, Qwen Code & more Comment Qwen to get Free promo code
English
@CodeCipherHQ @rohanpaul_ai I think we'll see even more in the coming years - the tech develops itself ⚡️
English
@rohanpaul_ai The craziest part of AI right now is how fast “almost frontier-level” has become normal 😳⚡ A year ago, matching GPT-level coding felt impossible for most open models.
English
@jsyqrt @rohanpaul_ai Actually we are finalizing our longread about Qwen 3.7 max - and we'll share it with you when live 😊
English
The 35-hour autonomous execution stat is the one that catches my attention. Running Markus agents for extended periods, the real challenge isn't whether the model can keep going — it's context drift. After 5-6 hours of continuous execution, even frontier models start making progressively worse decisions because earlier context gets compressed to near-uselessness. Would be curious what mechanisms they use to maintain coherence at 35h.
English
@CodeCipherHQ @rohanpaul_ai I would not consider that a bad thing! Technology is running at a great pace - and that opens quite a lot of opportunities for builders
English
@rohanpaul_ai Feels like we’re entering the era where model rankings change every 2 weeks 😭📈 Developers barely finish benchmarking one model before another suddenly jumps into the top 5.
English
@ShinkaIoT @DataChaz And it survives. We've had ran multiple tests - and it turned out to provide really consistent results + low hallucination rate even on a really long multi-tasks.
English
BOOM💥 Qwen3.7-Max drops on @AIMLapi and the stats are unreal.
> 92.4 GPQA Diamond
> Apex: 44.5
> 35+ hours of autonomous execution.
Let that sink in.
Go comment "Qwen" on their post, hit up their Discord, and claim your free access 👀
AI/ML API@aimlapi
Qwen3.7-Max on AI/ML API - built for the agent era GPQA Diamond (92.4), HMMT (97.1), Apex (44.5) Sustains 35+ hours of autonomous execution Works with Claude Code, Qwen Code & more Comment Qwen to get Free promo code
English
@aiseomastery from what we've had in our testing experiments - it almost never hallucinate / can hold context in a quite a consistent and long window
so absolutely impressive
English
@aimlapi 35+ hours of autonomous execution is the number that stands out, how does it hold up on complex multi step tasks?
English
