Are the models genuine? Are requests modified?
Are the models genuine? Are requests modified?
Which models are best for coding tasks?
Which models are best for coding tasks?
- deepseek-v4-pro — strong reasoning and multi-language code generation
- glm-5.1 — excellent instruction following and code completion
- kimi-k2.7-code — purpose-built for code with a large context window
- minimax-m2.5 — solid all-rounder for code review and explanation
Do you support vision or multimodal models?
Do you support vision or multimodal models?
content array of your message, following the standard OpenAI-compatible multimodal message format.What context window sizes are available?
What context window sizes are available?
| Model family | Max context |
|---|---|
| Kimi | 256K tokens |
| GLM | 203–205K tokens |
| MiniMax | 200K tokens |
| DeepSeek | 128K tokens |
Does my balance expire?
Does my balance expire?
What payment methods are accepted?
What payment methods are accepted?
- USDT (on-chain crypto)
- Stripe (card processing)
- PayPal
- Overseas credit and debit cards
Is there a monthly subscription fee?
Is there a monthly subscription fee?
How do I get free trial credits?
How do I get free trial credits?
Can users in mainland China use A2Agent?
Can users in mainland China use A2Agent?
What regions are supported?
What regions are supported?
How are my requests and data handled?
How are my requests and data handled?
Is my API key secure?
Is my API key secure?
- Never share it publicly or via chat
- Never commit it to source control (use environment variables instead)
- Revoke and regenerate immediately from the console if you suspect it has been exposed
I'm getting a 401 Unauthorized error
I'm getting a 401 Unauthorized error
I'm getting a 403 Forbidden error
I'm getting a 403 Forbidden error
403 response means your request was rejected after authentication. There are two common causes:- Suspended account — your account may have been suspended. Contact patchx2025@gmail.com to find out why and resolve it.
- Insufficient balance — your account balance has been exhausted. Top up your account from the console to resume making API calls.
403 is distinct from a 401: a 401 means the key itself is invalid or missing, while a 403 means the key is valid but the account cannot currently fulfill requests.I'm getting a 429 Too Many Requests error
I'm getting a 429 Too Many Requests error
429 response means you’ve exceeded the rate limit for your current tier. To resolve this:- Implement exponential back-off and retry logic in your client.
- Reduce request concurrency if you’re firing many parallel calls.
- Contact support at patchx2025@gmail.com or join the Discord to discuss higher rate-limit tiers.
A model is returning unexpected results
A model is returning unexpected results
- Verify the model ID is spelled exactly as shown on the Models Overview page — IDs are case-sensitive.
- Check that you’re not accidentally hitting a lower-priced group when you intend to use the default group.
- Review your system prompt and message structure for unintended instructions.
- If the issue persists, share the request details (minus sensitive content) in the Discord for community help.