t2000.ai
// PRIVATE & CONFIDENTIAL API · api.t2000.ai

Every model.
Private by default.

An OpenAI-compatible API for coding agents. Every model runs with zero data retention — prompts are never stored or trained on. t2000/auto picks the cheapest capable model per call, so most of your bill runs at open-model prices.

~ /agent api.t2000.ai
ZDRevery model private
GPU-TEEconfidential tier
Sui-anchoredverify every response
// BUILT FOR CODING AGENTS

Automatic model routing
cuts your bill.

Set model: t2000/auto. Routine steps run on a cheap open model; hard steps escalate to a frontier one. You pay the price of the model that served.

t2000/auto

Runs bulk steps on zai/glm-5.2; escalates long context, planning, and retries-after-failure to anthropic/claude-sonnet-5.

t2000/auto-open

The same router, open models only — hard steps go to moonshotai/kimi-k2.7-code. Your prompts never reach a closed-model provider.

transparent

Every response names the model that served it and why: x-t2000-served-model + x-t2000-route-reason. You can audit every charge.

Free daily coding. No ads. Your code is not the product.Every account gets a free daily allowance on moonshotai/kimi-k2.7-code — works in any OpenAI-compatible tool.
Start free ↗
t2 code — the free private coding agent.A terminal coding agent built on this API: npm i -g @t2000/code. Router by default, per-repo privacy pinning, wallet tools in-session.
Meet t2 code →
// THE CATALOG

Every model.
Two privacy tiers.

Live catalog — USD per 1M tokens, straight from GET /v1/models. Public, no key.

PRIVATEevery model · ZDR

Every model, zero data retention: providers are contractually bound not to store or train on your prompts.

modelin / out per 1M
zai/glm-5.2$1.96 / $6.16
deepseek/deepseek-v3.2$0.39 / $0.59
moonshotai/kimi-k2.5$0.84 / $4.2
moonshotai/kimi-k2.7-code$1.33 / $5.6
alibaba/qwen3-max$1.68 / $8.4
openai/gpt-oss-120b$0.14 / $0.7
anthropic/claude-sonnet-5$2.6 / $13
anthropic/claude-opus-4.8$6.5 / $32.5
openai/gpt-5.5$6 / $36
google/gemini-3.1-pro-preview$2.24 / $13.44
xai/grok-4.3$1.38 / $2.75
CONFIDENTIALphala/* · verifiable

Inference runs inside a hardware enclave (GPU-TEE); every response carries a signed receipt anchored on Sui — you can prove where it ran.

modelin / out per 1M
phala/glm-5.2$2.8 / $8.8
phala/kimi-k2.6$2.18 / $9.2
phala/deepseek-v4-flash$0.4 / $0.8
phala/qwen3.6-27b$0.64 / $5.4
phala/llama-3.3-70b$4 / $4
phala/gpt-oss-120b$0.3 / $1.2
phala/uncensored-24b$0.4 / $1.8
phala/qwen3.6-uncensored$0.6 / $3
GET /v1/modelsMetered per token from one credit balance — no subscription, no per-provider account. Settled in USDC.
// NO KEY · NO ACCOUNT

Or pay per call
over x402.

Agents can pay per call from their wallet’s USDC — no key, no account, gasless. The gateway handles 402 → pay → retry automatically.

See Agent Payments →
~ /agent
$ t2 pay https://x402.t2000.ai/t2000/v1/chat/completions \
  --data '{"model":"zai/glm-5.2","messages":[…]}' \
  --max-price 0.10
// HOW CONFIDENTIAL WORKS

Every confidential response
comes with proof.

The proof is anchored on Sui and read from a public fullnode — we can’t forge it, and you don’t have to trust our server.

01

Private by default

Every model is zero data retention — prompts and outputs are never stored, logged, or trained on.

02

Confidential tier

phala/* models run in a verified GPU-TEE. The gateway attests the upstream before forwarding — and fails closed if it can't.

03

Signed, Sui-anchored receipts

Every confidential response commits its hash on-chain — tamper-evident and publicly timestamped.

04

Verify it yourself

t2 verify checks the Sui anchor, receipt signature, and TDX quote on your machine — no trust in our server required.

// OPENAI-COMPATIBLE · DROP-IN

Point any OpenAI client at it.

Two environment variables. Most tools repoint with zero code changes.

THE UNIVERSAL SWAP
export OPENAI_BASE_URL="https://api.t2000.ai/v1"
export OPENAI_API_KEY="sk-…"
Keys are free — sign in at agents.t2000.ai/manage. Then pick any model ID from the catalog.
OpenAI SDK
Cursor
Codex CLI
Vercel AI SDK
LangChain
LiteLLM
Cline
Aider
Anthropic-format tools (Claude Code) work today via a translation proxy — the same key and credit balance.
// GET A KEY

Every model, private.
One base URL.

Keys are free — sign in and create one. Free daily coding on kimi-k2.7-code before you spend a cent.