🌋Comput3 $COM
28 October 2025 06:36
🚀 We just shipped support for Minimax M2 - the world's most powerful open source model for coding, tool calling & agents. It just launched today!
Most powerful GPUs (B200s) × most powerful agentic model × #x402 = compute scale 💪
Only
@
$COM 🌋
https://github.com/comput3ai/c3-vllm
Читать полностью…
🌋Comput3 $COM
26 October 2025 23:32
We're live at comput3.ai/chat! Chat with 22 models: Grok, Claude Sonnet/Opus, GPT-5, plus Hermes4 405B & Qwen3 Coder 480B. Discounted pricing on models we host. We now have #x402 via @. Use our UI, or just export the OpenAI API key. More models tomorrow! 🚀
Читать полностью…
🌋Comput3 $COM
26 October 2025 01:37
When we're quiet, we're just too busy cooking. x402 payments, all the models we recently added, all the partnerships we've locked in and even new GPU workloads like training on B200s. It's all on our updated website https://comput3.ai 🌋
Читать полностью…
🌋Comput3 $COM
24 October 2025 14:44
Engineering Breakthrough + Major Updates 🚀
We just pulled off something special. Through optimization and the sheer power of B200s, we're now running Hermes4 and Qwen3 Coder on the same hardware with FULL CONTEXT. No one else has B200s in production like this. Only comput3 can do this.
What This Means for You 💎
Qwen3-coder is back and effectively FREE for stakers. We're discounting our hosted models so heavily that premium stakers won't hit their budgets under normal use—we're treating you as power users. Other stakers get solid allocations too. Running 1000x coding agents? DM us and we'll work out epic custom pricing together.
Aggressive Pricing 💰
Our hosted models are currently 75% off. We're pushing toward 90% off. Goal: be at least 2x cheaper than anyone else for every model we run. These B200s are absolute monsters with tons of spare capacity right now, and our pricing reflects it. No one else can compete with this hardware.
launch.comput3.ai Revamp 🔧
We're retiring the underused ollama instances and freeing up capacity for what you actually want: media generation. Adding wan2.2, qwen-image, and other models that boot instantly. The new base images will be vastly superior and track upstream ComfyUI better.
What's Coming 🛠️
Docker images with SSH access and eventually custom Dockerfile support. A Python library (Modal-style) where a couple lines of code gets you Jupyter notebooks, VLLM instances, Whisper—whatever you need. We're making deployment dead simple.
Our Philosophy 🎯
External models remain pass-through at cost with zero markup. Want to pay in $SOL or $COM and stay completely anonymous while using Grok or Anthropic? That's why we offer it. Our mission is simple: host every major open source model at the best prices in the industry.
This is just the beginning. 🔥🌋
Читать полностью…
🌋Comput3 $COM
27 September 2025 19:35
We just dropped something massive for the community. Our exclusive partnership with Nous Research makes us the official hosting platform for Hermes4 - meaning you get direct access to one of the most advanced open-source models out there, straight from the source. This isn't just another integration; we're literally hosting their flagship model for the authors themselves and their users!
But here's the real kicker: we also launched subscriptions giving you access to all our favorite models. We now give you access to: hermes4:70b, hermes4:405b, deepseek-v3.1, kimi-k2, qwen3-coder:480b, qwen3-max, claude-sonnet-4, claude-opus-4.1, gpt-5, gpt-5-codex, grok-4, grok-code-fast-1, gemini-2.5-pro, gemini-2.5-flash!
Every single subscription automatically triggers COM token buybacks. Whether you pay with card, Solana, or stake - doesn't matter. Real revenue driving real demand for our token. No speculation, no promises - just pure utility creating consistent market pressure. We've built the first chat platform where growth actually benefits holders through automatic buybacks. This changes everything.
50% of every subscription goes to purchase $COM. You keep 50% of the $COM and the other 50% goes to combuybacks.sol, our community fund to pay referral fees and other community rewards. 🤝🌋
Signup now and try out all the models, all models are free to try within the chat interface for a limited time: https://console.comput3.ai
Full live stream is here on youtube: https://youtube.com/live/Jq6AiB1_qIs?feature=share
Читать полностью…
🌋Comput3 $COM
25 August 2025 11:44
👏 Thanks to everyone who participated 🌋
📢 Unfortunately, our current airdrop has concluded.
🚀 But don't worry! Subscribe to our channel to be the first to know about our next airdrop and exclusive opportunities:
👉 @
📱 /channel/comput3ai
💎 Stay tuned for more rewards and community events!
Читать полностью…
🌋Comput3 $COM
24 August 2025 16:19
Someone in our discord did this, we swear it wasn't us (wait until it finishes). $COM 🌋
Читать полностью…
🌋Comput3 $COM
23 August 2025 15:17
🔥 Thought we were just “web3”? Nah, we’re gunning for the big dogs.
Cursor, Anthropic, OpenAI—watch your backs.
Stake $COM or swipe with $SOL / $USDC / fiat.
Either way, your bag stays safe, your mind gets blown. 🌋
https://x.com/comput3ai/status/1959233588185997365
Читать полностью…
🌋Comput3 $COM
21 August 2025 20:08
Quickest way to test kimi-k2 and qwen3-coder running on comput3.ai directly from your browser! 🌋
Читать полностью…
🌋Comput3 $COM
21 August 2025 08:56
🚀 The word is out: we’re quietly becoming the hottest infrastructure layer in AI.
Kimi-K2 and Qwen3-Coder models are live, FP8-hosted, full context, B200-powered, and served outside China—exactly the combo the market was waiting for.
• Devs are already shipping at lightspeed—some can’t believe the integration is this easy.
• ComfyUI is dropping a fresh build that unlocks Qwen’s new image & editing models.
• GPU network is humming; next-gen LLMs are spinning up daily.
If your project needs compute or input/output tokens to power your AI apps, we’re taking those conversations now.
DMs are open, but discord is where the action is. Just get in touch with us.
https://discord.com/invite/DmaHAdGcNy
Читать полностью…
🌋Comput3 $COM
18 August 2025 14:29
Claude doesn't even know how true this is right now. 🌋
Читать полностью…
🌋Comput3 $COM
18 August 2025 09:51
Today's the big day, we just deployed kimi-k2 and qwen3-coder:480b to our api. If you want to play around with this, hop on our discord: https://discord.com/invite/DmaHAdGcNy
Alright @ it's the big day.
This is work in progress. We are doing a ton of stuff that you can't see, but it is up and we would appreciate testing and feedback. The inferencing API is now running on B200's. Right now everyone has access to all the models, we'll be closing this down as we go along.
We'll be launching paid subscriptions, we had a lot of discussions and honestly these models are so powerful that we'll probably be targeting $99/mo, but we'll start with $75/mo as a welcome offer. You won't have to use phantom, it'll be email and stripe (we'll do crypto too don't worry). We will be giving stakers access to what we consider reasonable limits (input/output tokens per month), but they'll be less than the subscriptions at least initially. Eventually we hope we can just map stakers to these plans, but for the moment we just added a bunch of 24/7 infra, which we're running in parallel to the GPU workloads most of you are familiar with.
The models on the api include the hermes3:70b and llama3:70b, both now running at 64K context (thanks B200s), but the big news is kimi-k2 and qwen3-coder. Right now everyone with an api key has access to these models. We'll be limiting them to stakers and subscribers later this week.
Keep in mind these are monstrous models and you might hit the Cloudflare's request timeouts after 100 seconds. That's not us - that's a cloudflare "feature". We're working on the api.compute3.ai (with an 'e') deployment as we speak, and that will be naked (no cloudflare, just a couple of load balancers) so it won't have that issue. We'll be adding instructions on how to set all of this up with claude code, qwen code and probably codex as well. In the mean time feel free to test, api.comput3.ai is now running a litellm-proxy so it should handle most of the shenanigans that the coding agents need. However, we're not routing /anthropic/* (or /openai/ for that matter) to our models yet, we're looking for a reasonable way to be able to leave you control over which goes where instead of picking one for you. I do believe overloading the /model should work, but I haven't tested this yet.
Enjoy and let us know how it goes. Here's some curls to get you started.
$ curl -X GET https://api.comput3.ai/v1/models | jq
{
"data": [
{
"id": "hermes3:70b",
"object": "model",
"created": 1677610602,
"owned_by": "openai"
},
{
"id": "kimi-k2",
"object": "model",
"created": 1677610602,
"owned_by": "openai"
},
{
"id": "qwen3-coder:480b",
"object": "model",
"created": 1677610602,
"owned_by": "openai"
},
{
"id": "llama3:70b",
"object": "model",
"created": 1677610602,
"owned_by": "openai"
}
],
"object": "list"
}
You can access these as a normal
/v1 API.
curl -X POST https://api.comput3.ai/v1/chat/completions \
-H "Authorization: Bearer c3_api_YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "kimi-k2",
"messages": [
{
"role": "system",
"content": "You are a helpful assistant."
},
{
"role": "user",
"content": "What is 2+2?"
}
]
}'
{"id":"chatcmpl-007e9d34e4474e78902934247b1f72bc","created":1755501320,"model":"kimi-k2","object":"chat.completion","system_fingerprint":null,"choices":[{"finish_reason":"stop","index":0,"message":{"content":"2 + 2 = 4","role":"assistant","tool_calls":null,"function_call":null},"provider_specific_fields":{"stop_reason":163586}}],"usage":{"completion_tokens":8,"prompt_tokens":24,"total_tokens":32,"completion_tokens_details":null,"prompt_tokens_details":null},"service_tier":null,"prompt_logprobs":null,"kv_transfer_params":null}
Читать полностью…
2231