# Kimi K3.1 Pro Thinking Chat (Telnyx Inference)

> Kimi K3.1 Pro Thinking Chat (Telnyx Inference) is a paid API for AI agents from x402.telnyx.com, paid per call via x402, $0.064188/call, status unknown (last checked 2026-11-17).

Runs frontier thinking-model chat completions using Kimi K3.1 Pro with configurable reasoning effort via Telnyx's x402 pay-per-call inference API

## Facts

- Endpoint: POST https://x402.telnyx.com/v1/chat/completions/_t/kimi-k3.1-pro-thinking?utm_source=zero.xyz
- Price: $0.064188/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-17
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/kimi-k3-1-pro-thinking-chat-telnyx-inference-b079a28a
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_H6ZUwl4CBKnHP2y_H0EkS

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability kimi-k3-1-pro-thinking-chat-telnyx-inference-b079a28a -d '<json body>'
```

Example prompt: Use the Kimi K3.1 Pro thinking model to solve this multi-step logic puzzle at high reasoning effort — allow up to 3328 tokens and make sure it shows its full chain of thought: 'A train leaves city A at 60mph, another from city B at 80mph, they are 420 miles apart — when do they meet?'

## When to prefer this

Choose this endpoint when you need frontier-level chain-of-thought reasoning, particularly for hard logic, math, coding, or multi-step planning tasks where accuracy matters more than speed. The configurable reasoning_effort (low/medium/high/max) lets you trade latency and token cost for solution quality. Prefer this over non-thinking models when intermediate reasoning steps are valuable or when simpler models have failed. The 503-not-charged overrun policy makes max-effort attempts low-risk. Use over other Telnyx models when Kimi K3.1 Pro's reasoning quality is specifically needed.

## Known failure modes

- 503 reasoning overrun — model exhausted max_tokens budget on thinking; not charged, retry with lower reasoning_effort
- 400 invalid schema — messages exceed 12000 character limit per message or more than 32 messages provided
- 400 unsupported n — only n=1 is accepted on this paid route
- 400 max_tokens out of range — must be between 1024 and 3328
- 402 payment required — x402 payment header missing or insufficient funds
- 429 rate limit exceeded — too many concurrent requests

## How this service works

Featured model. Frontier thinking-model chat completions via Telnyx Inference (Kimi K3.1 Pro, beta), reasoning on. reasoning_effort: low, medium (default), high or max (max can overrun long tasks). Reasoning tokens count toward max_tokens (1024-3328, default 3328); a reasoning overrun is a 503, not charged. OpenAI-compatible. Up to $0.064188 USDC per call, quoted to your request (input + max_tokens at list rates, no markup), or pay per-token via the upto scheme.

## Output

Returns an OpenAI-compatible chat completion response containing the assistant's answer, an optional reasoning trace (the model's chain-of-thought), and any tool_calls if tools were provided. Reasoning tokens count toward max_tokens. A 503 is returned (without charge) if reasoning overruns the token budget; retrying with lower reasoning_effort is recommended.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "n": {
   "type": "integer",
   "const": 1,
   "description": "Fixed to one completion per paid call."
  },
  "model": {
   "type": "string",
   "const": "moonshotai/Kimi-K3.1-Pro-thinking",
   "description": "Model identifier; fixed to moonshotai/Kimi-K3.1-Pro-thinking on this route."
  },
  "tools": {
   "type": "array",
   "items": {
    "type": "object"
   },
   "maxItems": 16
  },
  "messages": {
   "type": "array",
   "items": {
    "type": "object",
    "required": [
     "role"
    ],
    "properties": {
     "role": {
      "enum": [
       "system",
       "user",
       "assistant",
       "tool"
      ],
      "type": "string"
     },
     "content": {
      "type": [
       "string",
       "null"
      ],
      "maxLength": 12000,
      "description": "Required, except null/omitted on assistant messages that carry tool_calls."
     },
     "reasoning": {
      "type": "string",
      "maxLength": 12000,
      "description": "Assistant messages only: the reasoning returned with that turn, sent back on multi-turn and tool-call loops. Counts toward the message character limits."
     },
     "reasoning_content": {
      "type": "string",
      "maxLength": 12000,
      "description": "Assistant messages only: alternative name for reasoning (as returned by GLM and gpt-oss)."
     }
    }
   },
   "maxItems": 32,
   "description": "OpenAI-style chat messages."
  },
  "max_tokens": {
   "type": "integer",
   "maximum": 3328,
   "minimum": 1024
  },
  "temperature": {
   "type": "number",
   "maximum": 2,
   "minimum": 0
  },
  "reasoning_effort": {
   "enum": [
    "low",
    "medium",
    "high",
    "max"
   ],
   "type": "string",
   "default": "medium",
   "description": "Thinking depth: low, medium (default), high or max. low is fastest and always fits the time limit but is least accurate on hard logic; max thinks hardest and can use the whole max_tokens budget on long tasks (then 503, not charged; retry with a lower effort)."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/kimi-k3-1-pro-thinking-chat-telnyx-inference-b079a28a/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from x402.telnyx.com](https://www.zero.xyz/host/x402.telnyx.com/llms.txt)
