v1 · public beta

One key.
Every frontier model.

MyCoAI.dev routes any OpenAI- or Anthropic-compatible request to a healthy upstream and bills the token, not the seat. Drop in a base URL. Ship.

Start with $1.00 free→Read the docs
99.97% routing uptimeSSE token-accurate streamingUSD pay-as-you-go
MyCoAI.dev · routing inspector live · all regions
Requests · sec
187
p50 latency
312 ms
Routing uptime
99.97%
Models in service
38
active
§ 05 · Pricing

Up to 98% cheaper
than official pricing.

Same Claude and GPT-5.x models — at a fraction of list price. Every model shows exactly how much below official it lands. No tiers, no minimums, prompt-caching passes straight through.

See the catalogStart free
Example · GPT 5.5per 1M tokens
Official OpenAI
$5.00 / $30.00
You pay$0.50 / $2.98
−90% vs official
OpenAI SDK✦Codex CLI✦Anthropic SDK✦Claude Code✦LangChain✦Vercel AI SDK✦LlamaIndex✦OpenClaw✦litellm✦Continue.dev✦OpenAI SDK✦Codex CLI✦Anthropic SDK✦Claude Code✦LangChain✦Vercel AI SDK✦LlamaIndex✦OpenClaw✦litellm✦Continue.dev✦
§ 01 · Routing

One endpoint.
Many providers.

We sit between your code and the model's API. No prompt rewriting, no hidden system messages, no training on your data. We bill the token; the model does the rest.

  • →Auto-failover to a healthy channel within 800 ms.
  • →Translate Anthropic ↔ OpenAI on the fly.
  • →Per-key model whitelist, rate-limits, IP allowlist.
  • →Stream-flush SSE with byte-accurate token counting.
OGPT · 5.5
OGPT · 5.4
AClaude · Opus 4.7
AClaude · Sonnet 4.6
MMistral · L3
in service
§ 02 · Specifications

A relay engineered for production.

Latency
<200ms

Added over upstream. Streams flush as bytes arrive.

Streaming
SSE · WS

Token-accurate billing reconciled at stream close.

Failover
Multi-channel

Auto-route to a healthy upstream within 800 ms.

Translation
OAI ↔ Claude

Hit /v1/messages even when the channel speaks OpenAI.

§ 03 · Catalog

Models in service.

Full price list →
OpenAI22
  • GPT 6.1 Sol
    gpt-6.1-sol
    $0.20 / $0.99
    in / out · 1M
  • GPT 6 Sol
    gpt-6-sol
    $0.20 / $0.99
    in / out · 1M
  • GPT 6 Astra
    gpt-6-astra
    −90%
    $10.00 / $50.00
    $0.99 / $4.97
    in / out · 1M
  • GPT 6.1 Sol
    gpt-6.1-sol-pro
    $0.20 / $0.99
    in / out · 1M
Anthropic07
  • Claude Fable 5
    claude-fable-5
    $3.98 / $19.89
    in / out · 1M
  • Claude Haiku 4 5 20251001
    claude-haiku-4-5-20251001
    −60%
    $1.00 / $5.00
    $0.40 / $1.99
    in / out · 1M
  • Claude Opus 4 6
    claude-opus-4-6
    −60%
    $5.00 / $25.00
    $1.99 / $9.95
    in / out · 1M
  • Claude Opus 4 7
    claude-opus-4-7
    −60%
    $5.00 / $25.00
    $1.99 / $9.95
    in / out · 1M
§ 04 · Flow

From your curl to the model, in four steps.

01
01
Authenticate

Your sk-key is HMAC-checked at the edge. Rate-limit and IP rules enforced.

02
02
Pre-bill

We place a hold sized to your max_tokens. Insufficient balance fails fast.

03
03
Forward

Pick the highest-priority healthy channel for the requested model.

04
04
Reconcile

Stream the response; finalize the exact cost when the last token arrives.

Get a key.
Make a call. Today.

Sign-up takes a minute. We credit $1.00 to your wallet on email verification — that's a few thousand tokens to kick the tires.

Start with $1.00 freeDocumentation