# Mekler AI Batch LLM Completions > Mekler AI Batch LLM Completions is a paid API for AI agents from api.mekler.ai, paid per call via x402, $0.002796/call, status unknown (last checked 2026-10-18). Generate text or JSON completions for 1–20 independent tasks using GPT-6 Luna or GPT-6.1 Sol, with transparent pricing and async job recovery via x402 micropayment. ## Facts - Endpoint: POST https://api.mekler.ai/v1/compute/inference/generate-completions?utm_source=zero.xyz - Price: $0.002796/call - Payment: x402 - Status: unknown - Last checked: 2026-10-18 - Activations on Zero: 0 - Tags: x402 - Canonical page: https://www.zero.xyz/c/mekler-ai-batch-llm-completions-f4028c91 - Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_kYl0bGAzTIQ6R1hRytu9f 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 mekler-ai-batch-llm-completions-f4028c91 -d '' ``` Example prompt: Run these 3 tasks for me using auto model selection: summarize this article, classify this review as positive/negative/neutral, and translate this sentence to French — keep total spend under $0.05 and return JSON for the classification. ## When to prefer this Choose this endpoint when you need to run multiple independent LLM inference tasks in a single atomic call with upfront cost transparency, especially when you want to mix task types (summarize, classify, translate, code) or model tiers (Luna vs Sol) in one batch. Ideal for agentic workflows where result recovery and idempotency matter. Prefer over direct OpenAI API calls when you want x402 micropayment-based billing, no streaming overhead, and async job storage. Not suitable for tool-use, web browsing, real-time streaming, or tasks requiring more than 20 parallel completions. ## Known failure modes - max_price_usdc too low for the planned task cost — request rejected before payment - malformed messages array (missing role or content) — 400 validation error - task array exceeds 20 items — schema validation failure - model or task enum value not recognized — validation error - job still pending when polled — return recovery token and retry - content refusal by underlying model — that item is uncharged - output truncated at max_output_tokens — finish_reason is length - network or payment settlement failure — retry with original proof, do not repay ## How this service works Compare exact prices across eight models: GPT-6 Luna, GPT-6.1 Sol, Ministral 3B/8B/14B, Nemotron 3 Nano and Qwen3 Next/Coder Next. Review a recommendation and supported model/effort alternatives, then accept one quote or use automatic selection. Generate text or JSON with one USDC payment and no supplier account. Recover private results after completion. Recommendations use task rules, not a quality guarantee. No tools, browsing or streaming. ## Output Returns a job object with: the pre-execution plan (model chosen per task, reasoning, total price in micros), state (complete/pending), job ID, recovery token, blockchain receipt of the x402 payment, and per-task results including completion text, token usage (prompt/completion/total), finish reason, and the actual model used. Results are stored and recoverable via the original proof or recovery token. ## Request schema (JSON Schema) ```json { "type": "object", "properties": { "tasks": { "type": "array", "items": { "type": "object", "required": [ "messages" ], "properties": { "task": { "enum": [ "auto", "extract", "classify", "summarize", "translate", "write", "code", "analyze", "plan" ], "default": "auto" }, "model": { "enum": [ "auto", "gpt-6-luna", "gpt-6.1-sol", "ministral-3-3b", "ministral-3-8b", "ministral-3-14b", "nemotron-3-nano", "qwen3-next", "qwen3-coder-next", "gemma-4-e2b" ], "default": "auto" }, "contract": { "type": "object", "required": [ "kind" ], "properties": { "kind": { "enum": [ "extract_fields", "classify" ] }, "fields": { "type": "object", "maxProperties": 3, "minProperties": 1, "additionalProperties": { "type": "string" } }, "labels": { "type": "object", "maxProperties": 3, "minProperties": 2, "additionalProperties": { "type": "string" } } }, "additionalProperties": false }, "messages": { "type": "array", "items": { "type": "object", "required": [ "role", "content" ], "properties": { "role": { "enum": [ "system", "user", "assistant" ] }, "content": { "type": "string", "maxLength": 16000, "minLength": 1 } }, "additionalProperties": false }, "maxItems": 16, "minItems": 1 }, "priority": { "enum": [ "balanced", "economy", "quality", "speed" ], "default": "balanced" }, "response_format": { "enum": [ "text", "json_object" ], "default": "text" }, "max_output_tokens": { "type": "integer", "default": 1000, "maximum": 4096, "minimum": 32 } }, "additionalProperties": false }, "maxItems": 20, "minItems": 1 }, "max_price_usdc": { "type": "string", "pattern": "^(?:0(?:\\.[0-9]{1,6})?|1(?:\\.0{1,6})?)$", "maxLength": 8, "minLength": 1 }, "idempotency_key": { "type": "string", … (truncated) ``` ## Response schema (JSON Schema) ```json { "type": "json", "example": { "plan": { "models": [ { "index": 0, "model": "gpt-6-luna", "reason": "bounded_general_task" } ], "version": "inference-v4-purchase", "price_micros": 2796 }, "retry": "Recover with the original proof or private token. Do not pay again.", "state": "complete", "job_id": "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", "receipt": { "network": "eip155:8453", "finalized": true, "transaction": "0xcccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc", "price_micros": 2796 }, "results": [ { "index": 0, "model": "gpt-6-luna", "usage": { "total_tokens": 25, "prompt_tokens": 18, "completion_tokens": 7 }, "content": "The meeting starts at noon.", "finish_reason": "stop", "provider_model": "gpt-6-luna", "reported_model": "gpt-6-luna" } ], "expires_at": 1794528000, "recovery_token": "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb" } } ``` ## More - Live health (JSON, refreshed every minute): https://www.zero.xyz/c/mekler-ai-batch-llm-completions-f4028c91/health.json - [Zero catalog index](https://www.zero.xyz/llms.txt) - [Other services from api.mekler.ai](https://www.zero.xyz/host/api.mekler.ai/llms.txt)