500K token context
Half a million tokens in one conversation, enough for a large codebase or a long document set.
xAI released Grok 4.5 on July 8, 2026 as its flagship for coding, agentic tasks and knowledge work: a 500,000 token context window, text and image input, and the top score on agentic tool use in the independent Artificial Analysis index. On avots.ai you run it pay per use, with no X Premium or SuperGrok plan, and the same balance also covers Claude Opus 5, GPT-5.6, Gemini and 40+ image, video and music models.
Half a million tokens in one conversation, enough for a large codebase or a long document set.
Per million tokens, with cached input at 0.50 USD, so repeated context gets cheaper.
Graduate level science questions, with 83.3% on Terminal-Bench 2.1 and 64.7% on SWE-bench Pro.
Send screenshots, charts, diagrams and scanned pages, and get text back.
Reasoning effort low, medium or high, with high as the default, so you trade latency for depth.
Function calling and structured outputs work through the standard OpenAI compatible interface.
| Grok 4.5 | Grok 4.3 | Claude Opus 5 | |
|---|---|---|---|
| Price per 1M in / out | $2 / $6 | $1.25 / $2.50 | $5 / $25 |
| Context window | 500K tokens | 1M tokens | 1M tokens |
| Terminal-Bench 2.1 | 83.3% | not published | 89.1% |
| GPQA Diamond | 93.1% | not published | 93.2% |
| Vision | yes | yes | yes |
| On avots.ai | pay per use | pay per use | pay per use |
Grok 4.5 sits between the cheap Grok 4.3 and the frontier Claude Opus 5: it matches Opus 5 on graduate science at a fraction of the output price, and it leads on agentic tool use. Switch between all three in the picker, one balance covers them.
It topped the agentic tool use category of the Artificial Analysis index at launch, with an overall score of about 54.
83.3% on Terminal-Bench 2.1 and 64.7% on SWE-bench Pro, built for long shell and repository work.
Cached input at 0.50 USD per million tokens keeps repeated prompts and long agent loops affordable.
Set effort to low, medium or high per call and pay for depth only when the task needs it.
Sign up on app.avots.ai with email, Google or Telegram. Takes a minute.
1,000 tokens cost 0.99 EUR. No plan to choose, no renewal, the balance sits there until you use it.
Choose it in the model picker, or let Agent Avots route for you. Every reply shows its token cost.
app.avots.ai on desktop and mobile, no install needed.
Grok 4.5 in @AvotsAIbot, same balance, from any phone.
Use it inside Claude Desktop, Cursor or Cline via the MCP server.
The OpenAI compatible API takes your existing client, one key for every model.
1,000 tokens cost 0.99 EUR and a short Grok 4.5 answer costs a handful of them. The same balance unlocks Claude Opus 5, GPT-5.6, Gemini, Veo video and Nano Banana images, with no X Premium subscription.
Sign up and chat ✦Newer version. This model has a successor — same balance, same app.
Every LLM on avots.ai and which one to pick per task.
Anthropic's newest flagship, 1M context, 96% SWE-bench Verified.
OpenAI's newest family: Sol, Terra and Luna tiers, 1M context.
Moonshot's frontier open model, number one at coding, 1M context.
xAI's image and video model for fast visual generations.
What to use instead of ChatGPT Plus and how one balance replaces it.
avots.ai runs 40+ AI models for chat, image, video and audio on one balance. See the full list on the Tools page or open the app and start creating.