Unified inference routing for every frontier model.
Point your existing OpenAI SDK at api.nullhost.tech/v1 to route across Anthropic, OpenAI, xAI Grok, Google Gemini, DeepSeek, Meta, Mistral, and Qwen with 6-decimal token settlement.
from openai import OpenAI
client = OpenAI(
base_url="https://api.nullhost.tech/v1",
api_key="gh_live_9f8a7b6c5d4e..."
)
stream = client.chat.completions.create(
model="claude-3-7-sonnet",
messages=[{"role": "user", "content": "Analyze system latency."}],
stream=True
)Routing verified via NullHost Edge. Hybrid reasoning and 65,536-token long-form continuation active.
Every frontier AI lab. Authentic provider icons & unified billing.
Transparent per-token pricing across all providers
Zero monthly fees. Every new account starts with $5.00 in free credit.
| Model | Provider | Specialty | Context | Prompt / 1M | Completion / 1M | Action |
|---|---|---|---|---|---|---|
Claude 3.7 Sonnet claude-3-7-sonnet | Anthropic | Flagship | 200K | $3.00 | $15.00 | Try |
GPT-4o gpt-4o | OpenAI | Multimodal | 128K | $2.50 | $10.00 | Try |
OpenAI o3-mini o3-mini | OpenAI | Reasoning | 200K | $1.10 | $4.40 | Try |
Grok 2 grok-2 | xAI | xAI Flagship | 128K | $2.00 | $10.00 | Try |
Grok 2 Vision grok-2-vision | xAI | Vision | 32K | $2.00 | $10.00 | Try |
DeepSeek R1 deepseek-reasoner | DeepSeek | Reasoning | 64K | $0.55 | $2.19 | Try |
Gemini 2.0 Flash gemini-2.0-flash | 1M Context | 1M | $0.10 | $0.40 | Try | |
Llama 3.3 70B Instruct llama-3.3-70b-instruct | Meta AI | Open Weights | 128K | $0.40 | $0.80 | Try |
Mistral Large 2 mistral-large-latest | Mistral AI | Enterprise | 128K | $2.00 | $6.00 | Try |
Codestral codestral-latest | Mistral AI | Coding | 256K | $0.30 | $0.90 | Try |
Qwen 2.5 72B Instruct qwen-2.5-72b-instruct | Alibaba Qwen | 72B Dense | 128K | $0.35 | $0.40 | Try |
Perplexity Sonar Pro sonar-pro | Perplexity | Live Web | 200K | $3.00 | $15.00 | Try |
Engineered for predictable production workloads
Switch models with a single parameter, inspect every token charge, and stream long-form completions without timeout drops.
OpenAI-Compatible Contract
Keep your existing OpenAI SDK, LangChain, Vercel AI SDK, Cursor, or cURL calls. Change only the base URL and API key.
65,536-Token Continuation
Automatic server-side continuation stitched over SSE so large codebases and multi-file generations never truncate mid-stream.
6-Decimal Token Ledger
Every request settles exact prompt and completion tokens to $0.000001 precision with unused pre-authorization holds released immediately.
Isolated Concurrency Pools
Dedicated upstream channels per model family prevent traffic spikes on one provider from degrading latency on others.
SHA-256 Key Security
Customer API keys are hashed at rest using SHA-256 with configurable per-key RPM rate limits and instant one-click revocation.
Mobile & Desktop Workbench
Compare responses across all 55 models with syntax-highlighted code blocks, raw markdown toggles, and full mobile navigation.
Ready to route your first request?
Create an account to receive $5.00 in complimentary gateway credit and mint your first API key in seconds.