PRICING
The most cost-efficient AI infrastructure, with full control over every request.
Save up to $50/mo
For indie devs and early-stage teams validating AI features in production — light traffic, real users.
Discount applied automatically.
Start SavingSave up to $1,000/mo
For growing teams shipping AI to a real user base — steady traffic, multiple features, predictable spend.
Discount applied automatically.
Start SavingSave up to $15,000/mo
For high-volume AI products at scale — multiple workloads, heavy concurrency, strict reliability.
Discount applied automatically.
Start SavingFor developers exploring models and testing ideas before committing to a stack.
$1.00/day billed monthly
Start BuildingX4 tokens
For developers actively integrating AI into real projects and shipping regularly.
$3.33/day billed monthly
Start BuildingX2 daily limit
For professional developers running AI as a daily tool across multiple projects at once.
$6.67/day billed monthly
Start Building
Each plan converts your monthly fee into a token pool you can spend across any model.
Here’s what that pool gets you on the models people actually use to ship.
Up to x10
Token value per $1 spent
Lite
$30/mo
Pro
$100/mo
Plus
$200/mo
Claude Opus 4.8
Anthropic · $25 / 1M tokens
6.9 М
~$173
23.1 М
~$578
46.2 М
~$1,155
Claude Sonnet 4.6
Anthropic · $15 / 1M tokens
11.5 М
~$173
38.5 М
~$578
76.9 М
~$1,154
GPT-5.5
OpenAI · $30 / 1M tokens
5.8 М
~$174
19.4 М
~$582
38.7 М
~$1,161
GPT-5.3 Codex
OpenAI · $14 / 1M tokens
12.5 М
~$175
41.8 М
~$585
83.6 М
~$1,170
Gemini 3.5 Flash
Google · $9 / 1M tokens
19.4 М
~$175
64.5 М
~$581
129 М
~$1,161
Grok 4.3
xAI · $3 / 1M
100 M
~$250
334 M
~$833
667 M
~$1,667
DeepSeek V4 Pro
DeepSeek · $1 / 1M
300 M
~$261
1 B
~$870
2 B
~$1,740
DeepSeek V4 Flash
DeepSeek · $0.28 / 1M
1.1 B
~$300
3.6 B
~$1,000
7.1 B
~$2,000
Kimi K2.6
Moonshot · $4 / 1M
50 M
~$200
167 M
~$667
334 M
~$1,333
Maximum retail value if you spend the whole subscription on one model.
up to ~$300
you pay $30
up to ~$1,000
you pay $100
up to ~$2,000
you pay $200
For every $1 spent, you’re getting up to $10 in token costs across the most powerful models.
Enterprise
Inference
Wholesale rates passed through, no caps.
Deployment
VPC, on-prem, or your provider keys.
Support
Private Slack channel, priority response.
Contract
Multi-entity invoicing, custom terms.
Claude
Anthropic: Opus, Sonnet, Haiku
ChatGPT
OpenAI: GPT-5, O-Series, Codex
Gemini
Google: Pro, Flash, Ultra
ElevenLabs
Voice: TTS, Conversational
Open-source models
DeepSeek, Qwen, GLM, MiniMax, Kimi
Everything included
The toggle changes the billing model. The product doesn’t change. Every plan ships with the full LLM.API surface.
Forget about downtime with LLM API smart routing, which guarantees you a reliable workspace
Swap your API key, keep your code. Works with every OpenAI SDK and tool on day one.
Rule-based router, automatic fallback on provider outages, reserved throughput when you need it.
Test prompts side-by-side across models, version them, and push to production with one click.
Per-key IAM rules, per-team budgets, hard rate limits. Nothing surprises your finance lead.
Prompts and responses are never stored. Default-on for every plan, no extra configuration.
Bring your own provider keys, deploy to VPC or on-prem, pick the region your data lives in.
Audited compliance documentation, BAAs, DPAs and security questionnaires ready when you ask.
Request-level traces, model-mix breakdowns, exportable usage data, alerts when patterns shift.
Token pricing
Built to the standards your security, legal, and compliance teams already trust.
FAQ
AI for Production is pay-as-you-go: you top up credits, get an automatic volume discount (10–20%) on every token, and run with no rate or usage limits. AI for Coding is a flat monthly subscription ($30–$200) with a fixed list of included models and 5-hour, daily, and monthly token limits.
The Coding subscription covers a curated set of frontier models: Claude Opus 4.8, Claude Sonnet 4.6, Claude Haiku 4.5, GPT-5.5, GPT-5.3 Codex, Gemini 3.5 Flash, Grok 4.3, DeepSeek-V4-Pro, DeepSeek-V4-Flash, and Kimi K2.6. Your monthly fee converts into a token pool you spend across these, and the same dollar stretches further on the cheaper ones (up to ~6x token value per $1).
Your tier (Build/Ship/Scale) is set automatically by your last 30 days of usage and re-evaluated every billing cycle — there are no contracts or negotiation. Just add credits and the discount applies on top-up.
No. Credits never expire, so unused balance carries forward indefinitely.
No. Every plan — and both billing tracks — ships the full LLM.API surface: smart routing, fallback, EvalLab, budget controls, BYOK, zero data retention, and SOC 2 / GDPR / CCPA / ISO compliance. The toggle only changes how you’re billed.
It’s drop-in OpenAI-compatible — swap your API key and keep your code, since it works with every OpenAI SDK and tool on day one.
