Buy credits once.
Spend on every AI model.
Top up once
Pick a credit pack from $10 to $1,000+. Pay in seconds with card, ACH, or crypto.
Get the multiplier
Every dollar buys you up to 10× the equivalent provider credit. Locked in at top-up.
Drop in our key
Swap one base URL. Your existing OpenAI / Anthropic SDK code keeps working — point it at neokens.
Spend across 16+ models
GPT-5.x Codex/Flagship, Claude 4.5/4.6, Gemini 2.5/3.x — one key, one balance, no per-vendor accounts.
One balance.
Every model worth using.
Prices below are USD per 1M input tokens. Output and image rates are listed in the docs. All models share the same API.
Claude Opus 4.7
The absolute peak of AI reasoning. For the most complex challenges.
Claude Sonnet 4.6
Exceptional performance for large-scale enterprise development.
Claude Sonnet 4.5
The balanced choice for rapid feature development and debugging.
Claude Haiku 4.5
Ultra-fast, efficient responses for high-volume automated tasks.
Gemini 3.1 Pro
State-of-the-art multimodal reasoning with 1M+ context window.
Gemini 2.5 Pro
Exceptional coding and complex problem solving performance.
Gemini 3.1 Pro Low
Efficient Pro-tier for cost-sensitive and high-volume workloads.
Gemini 3 Flash
High-speed inference with reliable quality for daily tasks.
GPT-5.5
Pinnacle reasoning, multimodal mastery, unmatched on every benchmark.
GPT-5.4
Next-level reasoning and superior performance across all tasks.
GPT-5.1
Next-gen flagship with unparalleled reasoning and multimodal capabilities.
GPT-5.1 Codex
Code-specialized variant. Purpose-built for software engineering.
GPT-5.2 Codex
The most capable code model. Ideal for complex refactors and architecture.
GPT-5.3 Codex
Ultra-fast code generation for real-time development workflows.
$1 in.
$10 out.
We aggregate enterprise commitments across all major providers and pass the volume discount back to you. Drag the slider — see exactly what your top-up unlocks.
One base URL.
Zero rewrites.
If your code already calls OpenAI, swap the base URL and you're done. Anthropic, Google, Llama — same shape, same SDK, one bill.
Pay once. Run thousands of completions.
- 10× multiplier
- All 9+ models
- Email support
- Credits never expire
- Everything in Starter
- Priority routing
- Slack support
- Usage analytics
- Spend alerts
- Volume bonuses
- Dedicated capacity
- Invoice billing
- Solutions engineer
- 99.99% SLA