Official models · up to 90% off

The LLM router built for
Claude Code and Codex

https://api.agentapiai.com/v1
BASE URL
MODEL PROVIDERS
CODING AGENTS & CLIENTS
Claude Code
Codex
CC Switch
OpenClaw
Works with 10+ coding agents and clients
  • Smart Routing
Claude
GPT
Gemini
  • Smart Routing
  • Prompt Caching
  • Auto Fallbacks
  • One Bill
  • Spend Tracking
  • Never Downgraded

Trusted by 100,000+ dev teams and indie builders

Live pricing · up to 90% off

Prices update in real time and move with upstream costs. Each request is billed at the discount in effect when it's made. All prices in USD per 1M tokens.See live discounts

Model Context Input (List) Output (List) Input (TeamoRouter) Output (TeamoRouter) Provider Uptime (SLA)
Claude Opus 5.5AnthropicContext — — $4Cache $0.2$20 $1.144Cache $0.0572$5.72 —
Claude sonnet 5.5AnthropicContext 1M 1M $2Cache $0.1$10 $0.578Cache $0.0289$2.89 —
GPT-6 AstraOpenAIContext 1M 1M $10Cache $1$50 $2.14Cache $0.214$10.7 —
GPT 6.1 SolOpenAIContext 1M 1M $2Cache $0.1$10 $0.428Cache $0.0214$2.14 —
Gemini 4 ArgonGoogleContext — — $2Cache $0.1$10 $0.61Cache $0.0305$3.05 —
GPT Image 2.5 SunburstOpenAIContext — — $5Cache $1.25$30 $0.5Cache $0.125$3 —
Claude Fable 5.1AnthropicContext 1M 1M $10Cache $0.25$50 $2.98Cache $0.0745$14.9 —
Gemini 3.8 FlashGoogleContext 1M 1M $0.75Cache $0.075$3.75 $0.2175Cache $0.0218$1.0875 —
DeepSeek V4.1 FlashDeepSeekContext 1M 1M $0.15Cache $0.003$0.6 $0.1275Cache $0.0026$0.51 —
DeepSeek V4.1 Flash福利版DeepSeekContext 1M 1M $0.15Cache $0.003$0.6 $0.0002Cache $0$0.0006 —
Deepseek V4.1 Flash 1017DeepSeekContext — — $0.15Cache $0.003$0.6 $0.1275Cache $0.0026$0.51 —
GLM-5.3-Flash (Ox Alpha)GLMContext 1M 1M $0.114Cache $0.03$0.4 $0.057Cache $0.015$0.2 —
GLM 5.3 Flash福利版GLMContext 1M 1M $0.114Cache $0.03$0.4 $0.0001Cache $0$0.0004 —
Grok 4.6XaiContext 500K 500K $2Cache $0.5$6 $0.2Cache $0.05$0.6 —
GPT Image 2OpenAIContext — — $5Cache $1.25$30 $0.5Cache $0.125$3 —
GPT Image 2.5 FlareOpenAIContext — — $5Cache $1.25$30 $0.5Cache $0.125$3 —
GPT-5.6 SolOpenAIContext 1M 1M $5Cache $0.5$30 $1.07Cache $0.107$6.42 —
GPT-5.5OpenAIContext 1M 1M $5Cache $0.5$30 $3Cache $0.3$18 —
GPT-5.4OpenAIContext 1M 1M $2.5Cache $0.25$15 $2.06Cache $0.206$12.36 —
GPT-5.6 TerraOpenAIContext 1M 1M $2Cache $0.2$12 $0.342Cache $0.0342$2.052 —
GPT-6 SolOpenAIContext 1M 1M $2Cache $0.2$10 $0.428Cache $0.0428$2.14 —
GPT 6.1 TerraOpenAIContext — — $1Cache $0.05$5 $0.171Cache $0.0086$0.855 —
GPT-5.4-MiniOpenAIContext 1M 1M $0.75Cache $0.075$4.5 $0.618Cache $0.0618$3.708 —
GPT-5.6 LunaOpenAIContext 1M 1M $0.2Cache $0.02$1.2 $0.12Cache $0.012$0.72 —
GPT-6 LunaOpenAIContext 1M 1M $0.1Cache $0.01$0.5 $0.05Cache $0.005$0.25 —
GPT Image 2.6 SunburstOpenAIContext — — $0$0 $0$0 —
GPT Image 2.6 Sunburst 4K UpscaleOpenAIContext — — $0$0 $0$0 —
GPT Image 2.5 Sunburst 4K UpscaleOpenAIContext — — $0$0 $0$0 —
GPT Image 2.6 FlareOpenAIContext — — $0$0 $0$0 —
GPT Image 2.6 Flare 4K UpscaleOpenAIContext — — $0$0 $0$0 —
GPT Image 2.5 Flare 4K UpscaleOpenAIContext — — $0$0 $0$0 —
GPT Image 2 4K UpscaleOpenAIContext — — $0$0 $0$0 —
Claude Fable 5AnthropicContext 1M 1M $10Cache $1$50 $2.96Cache $0.296$14.8 —
Claude Opus 4.8AnthropicContext 1M 1M $5Cache $0.5$25 $1.445Cache $0.1445$7.225 —
Claude Opus 4.6AnthropicContext 1M 1M $5Cache $0.5$25 $1.44Cache $0.144$7.2 —
Claude Opus 5AnthropicContext 1M 1M $5Cache $0.5$25 $1.43Cache $0.143$7.15 —
Claude Opus 4.7AnthropicContext 1M 1M $5Cache $0.5$25 $1.44Cache $0.144$7.2 —
Claude Sonnet 4.6AnthropicContext 1M 1M $3Cache $0.3$15 $0.861Cache $0.0861$4.305 —
Claude Sonnet 5AnthropicContext 1M 1M $2Cache $0.2$10 $0.578Cache $0.0578$2.89 —
Claude Haiku 4.5 (20251001)AnthropicContext — — $1Cache $0.1$5 $0.272Cache $0.0272$1.36 —
Claude haiku 5.5AnthropicContext — — $0.1Cache $0.01$0.5 $0.0288Cache $0.0029$0.144 —
Gemini 3 Pro ImageGoogleContext — — $2$120 $1$60 —
Gemini 3.1 Flash ImageGoogleContext — — $0.5$60 $0.25$30 —
Nano Banana 2.1GoogleContext — — $1.5$30 $0.75$15 —
Gemini 3.1 Pro PreviewGoogleContext 2M 2M $2Cache $0.2$12 $0.884Cache $0.0884$5.304 —
Gemini 3.5 FlashGoogleContext 1M 1M $1.5Cache $0.15$9 $0.5145Cache $0.0515$3.087 —
Gemini 3.6 FlashGoogleContext 1M 1M $0.75Cache $0.075$3.75 $0.3038Cache $0.0304$1.5188 —
Gemini 3.7 FlashGoogleContext 1M 1M $0.75Cache $0.075$3.75 $0.231Cache $0.0231$1.155 —
Gemini 3.5 Flash LiteGoogleContext 1M 1M $0.3Cache $0.03$2.5 $0.087Cache $0.0087$0.725 —
DeepSeek V4 Pro 260425DeepSeekContext — — $1.74Cache $0.145$3.48 $1.479Cache $0.1233$2.958 —
DeepSeek V4 Pro 正式版DeepSeekContext 1M 1M $0.66Cache $0.022$1.98 $0.561Cache $0.0187$1.683 —
DeepSeek V4 FlashDeepSeekContext 1M 1M $0.44Cache $0.014$1.32 $0.374Cache $0.0119$1.122 —
DeepSeek V4 Flash福利版DeepSeekContext 1M 1M $0.44Cache $0.014$1.32 $0.0004Cache $0$0.0013 —
DeepSeek V4 Flash Vision ExpDeepSeekContext 1M 1M $0.15Cache $0.003$0.6 $0.15Cache $0.003$0.6 —
Kimi K3MoonshotContext — — $3Cache $0.3$15 $1.5Cache $0.15$7.5 —
Kimi K3.1 ProMoonshotContext — — $2Cache $0.2$10 $1Cache $0.1$5 —
Grok 4.7XaiContext — — $2Cache $0.5$6 $0.242Cache $0.0605$0.726 —
Jev 1.14.0TypeSafeContext — — $0.042$0 $0.042$0 —
Jev 1.13.0TypeSafeContext — — $0.042$0 $0.042$0 —
Qwen 3.7 MaxQwenContext — — $1.7143Cache $0.3429$5.1429 $0.8572Cache $0.1715$2.5715 —
Qwen3.8 MaxQwenContext — — $1.7143Cache $0.2143$5.1429 $0.8572Cache $0.1072$2.5715 —
Qwen 3.8 FlashQwenContext — — $0.1143Cache $0.0143$0.3857 $0.0686Cache $0.0086$0.2314 —
Qwen3.7 FlashQwenContext — — $0.0286Cache $0.0057$0.1143 $0.0286Cache $0.0057$0.1143 —
GLM-5.3GLMContext 1M 1M $1.4Cache $0.26$4.4 $0.7Cache $0.13$2.2 —
GLM-5.2GLMContext 1M 1M $1.4Cache $0.26$4.4 $0.7Cache $0.13$2.2 —
Doubao Seed 2.1 ProByteDanceContext — — $0.8571Cache $0.1714$4.2857 $0.4286Cache $0.0857$2.1429 —
Doubao Seed 2.0 LiteByteDanceContext — — $0.0857Cache $0.0171$0.5143 $0.0857Cache $0.0171$0.5143 —
Doubao Seed 2.1 LiteByteDanceContext — — $0.1143Cache $0.0229$0.3857 $0.0571Cache $0.0114$0.1929 —
Seedance 2.0 R2VByteDanceContext — — $0$0 $0$0 —
Seedance 2.5 Flex R2VByteDanceContext — — $0$0 $0$0 —
Seedance 2.5 FlexByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 Mini R2VByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 MiniByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 MiniByteDanceContext — — $0$0 $0$0 —
Seedance 2.5ByteDanceContext — — $0$0 $0$0 —
Seedance 2.5 R2VByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 FastByteDanceContext — — $0$0 $0$0 —
Seedance 2.5ByteDanceContext — — $0$0 $0$0 —
Seedance 2.0ByteDanceContext — — $0$0 $0$0 —
Seedance 2.0ByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 FastByteDanceContext — — $0$0 $0$0 —
Seedance 2.0 Fast R2VByteDanceContext — — $0$0 $0$0 —
MiniMax M3MiniMaxContext — — $0.3Cache $0.06$1.2 $0.3Cache $0.06$1.2 —
MiniMax M3.1 FlashMiniMaxContext — — $0.15Cache $0.015$0.6 $0.075Cache $0.0075$0.3 —
MiniMax H3 MaxMiniMaxContext — — $0$0 $0$0 —
MiniMax H3MiniMaxContext — — $0$0 $0$0 —
MiniMax H3 MaxMiniMaxContext — — $0$0 $0$0 —
MiniMax H3 Max R2VMiniMaxContext — — $0$0 $0$0 —
MiniMax H3 R2VMiniMaxContext — — $0$0 $0$0 —
MiniMax H3MiniMaxContext — — $0$0 $0$0 —
MiMo V2.6 ProXiaomiContext — — $0.4286Cache $0.0036$0.8571 $0.2143Cache $0.0018$0.4286 —
MiMo V2.6 FlashXiaomiContext — — $0.1429Cache $0.0029$0.2857 $0.0715Cache $0.0015$0.1429 —
Claude Haiku 4.5AnthropicContext 200K 200K $1Cache $0.1$5 $0.244Cache $0.0244$1.22 —
Enterprise customer?Get enterprise pricing and exclusive discounts
139B+tokens routed yesterday
>99%prompt cache hit rate
99.98%routing SLA
Hoursto new-model support

Pay Less for Every Token

Pay only for what you use. Model rates start at just 10% of the official price — discounts applied automatically, per model.

Credits never expire · Pay by card or USDT

Quick start

Point your existing tools at TeamoRouter. A one-line change — official SDKs, standard endpoints, nothing to relearn.

View API docs
main.py

Point the base URL at TeamoRouter, drop in your API key — the rest of your code stays the same.

Running in Production

All three numbers come from live production traffic — the same data behind the pricing table and leaderboard on this page.

PRICE 90%off

2,000+ providers compete on price in real time under continuous quality monitoring, so every request gets the best available rate. Live pricing for every model is published on this page.

TIME TO FIRST TOKEN 2.4s

Latency and throughput on par with going direct to the provider, measured across all production traffic. Slow routes are pulled from rotation within minutes; the live numbers are on this page.

UPTIME 99.98%

Our Harness Routing engine backs a production-grade SLA and cache-hit guarantee: automatic failover across providers, around the clock, no downtime.

A Router You Can Build Your Business On

Traditional gateways weren't built for agent workloads. TeamoRouter is tuned for Claude Code, Codex, and long agentic sessions — without sacrificing the basics.

Model Integrity Get exactly the model you request. Every request is routed to the model and protocol you select through vetted upstream providers, never silently swapped or downgraded.
Data Privacy Prompts and completions are processed only for routing, metering, billing, abuse prevention, and support. Logs are never used to train models or sold as usage data.
Spend Tracking Billing details show every request, including model, token usage, applied discount, and final charge. Teams can see exactly where credits are spent instead of guessing from aggregate spend.

How are the prices this low?

One account, one bill. We vet providers, so you don't have to.

Providers compete for your traffic Model providers compete for every request; only those that pass our quality gates are eligible. You get market rates without doing the price discovery yourself.
Buy at scale We secure enterprise-scale volume commitments with vetted model providers. That is where the discount comes from.
Investor-backed pricing Backed by leading funds, we're passing part of that on as price subsidies — a limited-time boost on top of volume discounts.

Production-grade routing

Routes that pass strict quality validation and latency testing — prioritizing availability, response speed and result consistency.

  • PriceUp to 90% off
  • LatencyFirst-token latency on par with official APIs
  • Best forProduction workloads · critical paths · quality-sensitive work
  • AssuranceStrict validation · continuous live monitoring & replacement

Live discounts

Discounts are repriced every hour as upstream costs move. Each request is billed at the discount in effect when it's made.

No matching models Try another keyword, or tap All to clear filters
NOW—
AVG—
BEST—
NET INPUT—
NET OUTPUT—
NET CACHE—

48 hours · hourly · dashed = Average

Usage leaderboard

Updated daily at 00:00 UTC+8

Daily token volume

Total tokens routed through TeamoRouter per day

354.3Btokens ↗ 11.0% vs yesterday

Top agents

Ranked by tokens routed this week · change vs last week

  • 1. Codex Desktop9817.1K requests 1170.1Btokens↗ 30.2%
  • 2. Claude CLI1793.1K requests 268.2Btokens↗ 75.1%
  • 3. VSCode863.8K requests 107.4Btokens↗ 45.4%
  • 4. Claude Desktop517.5K requests 56.9Btokens↗ 104.3%
  • 5. OpenClaw189.1K requests 19.9Btokens↗ 151.2%

Top models

Ranked by tokens routed this week · change vs last week

  • 1. GPT-5.6 SolTTFT 1908ms · 99.46% success 1068.3Btokens↗ 31.8%
  • 2. GPT-5.6 TerraTTFT 2133ms · 98.56% success 251.6Btokens↗ 99.5%
  • 3. GPT-5.5TTFT 1562ms · 98.70% success 136.8Btokens↘ 19.3%
  • 4. Claude Opus 4.8TTFT 2371ms · 99.35% success 109.1Btokens↗ 122.6%
  • 5. GPT-5.6 LunaTTFT 2074ms · 98.93% success 97.7Btokens↗ 57.6%

Updated daily at 00:00 UTC+8 · Last updated

Cut Your AI Spend up to 90%

Top up in the dashboard and usage is drawn down per request — so the budget you set goes further.

Credits never expire · Pay by card or USDT

FAQ

A single API that sits between your tools and model providers. You call one endpoint with one key; the router sends each request to Claude, GPT, Gemini or others, meters usage, and gives you one bill.

Yes — both formats. The OpenAI-compatible endpoint works with official OpenAI SDKs and Codex by changing the base URL. Claude Code connects through the Anthropic-compatible endpoint with two environment variables.

Volume. We commit to enterprise-scale usage with vetted providers and pass the difference through. Discounts vary per model and move with upstream costs — the pricing table above is live.

No. Prompts and completions are processed only for routing, metering, billing, abuse prevention and support. They are never used for training and never sold.

No. Buy credits when you want, use them whenever. No subscription required.

TeamoRouter

Every Request Takes The Best Route

TeamoRouter - AI API Gateway for Claude Code & Codex