DeepInfra vs OpenRouter
DeepInfra vs OpenRouter
DeepInfra is a host: it runs open models on its own GPUs at its own prices. OpenRouter is a router: one API in front of many hosts, DeepInfra among them, with a fee when you buy credits. FullsendAI is a third option: one key with a web chat and one-command setup for your coding agents, no top-up fee and every model in the USA.
Which one fits
Choose DeepInfra if…
- You use open models and want to pay the host's own price, with no router in between.
- You want batch (20% off) and Flex tiers, speech or GPUs.
Choose OpenRouter if…
- You want 456 models behind one key, from many providers.
- You want to steer requests between providers, with fallbacks and precision filters.
Choose FullsendAI if…
- You want one balance for the API, a web chat and your coding agents.
- You want no fee on top-ups, by card or crypto (OpenRouter: 5.5% by card ($0.80 minimum) and 5% in crypto).
- You need every request served in the USA.
At a glance
Each DeepInfra and OpenRouter cell links to its source and the date we checked it.
What it is
Models
Markup on model prices
Fee on card top-ups
Ways to pay
OpenAI-compatible API
Anthropic Messages API
Claude, GPT and Gemini
Routing and failover
Where inference runs
Prompt logging by default
Web chat
Desktop coding agent
Own tuned models
| Feature | FullsendAI | DeepInfra | OpenRouter |
|---|---|---|---|
| What it is | One API key, a web chat and one-command setup for coding agents | Inference host on its own GPUs; also GPU rentalSource · | Router in front of many model providersSource · |
| Models | 4 models | 173 models on its OpenAI-compatible endpointSource · | 456 modelsSource · |
| Markup on model prices | 0% on open models | Sets its own list pricesSource · | None on inference, per its FAQSource · |
| Fee on card top-ups | None | None statedSource · | 5.5% when you buy credits, $0.80 minimumSource · |
| Ways to pay | Card and crypto (USDC, USDT and more) | Card, prepaid balance or invoice; crypto not listedSource · | Card and cryptoSource · |
| OpenAI-compatible API | Yes (Chat Completions) | YesSource · | YesSource · |
| Anthropic Messages API | Yes; works with Claude Code | Yes; documents Claude Code setupSource · | Yes; documents Claude Code setupSource · |
| Claude, GPT and Gemini | Not yet | Claude and Gemini; no GPTSource · | YesSource · |
| Routing and failover | We pick the host, with automatic failover; no provider controls | Single host; Priority (1.5×) and Flex (0.8×) tiersSource · | Provider ordering, fallbacks and filters you controlSource · |
| Where inference runs | United States, every model | United StatesSource · | Many providers; the route decides the regionSource · |
| Prompt logging by default | API prompts not stored unless you turn on 30-day body storage | Not stored or logged; Google and Anthropic models follow their own termsSource · | Prompts not logged by default; opt in for a 1% discountSource · |
| Web chat | Yes | YesSource · | Yes (chatroom)Source · |
| Desktop coding agent | Not yet: one command sets up Claude Code, Codex and more | No desktop agent; Claude Code setup guideSource · | No desktop agent; setup guides for third-party agentsSource · |
| Own tuned models | Nef 4.0 and Sifu 3.5, tuned for security research and code | NoSource · | No; routers such as openrouter/autoSource · |
Fees
What each charges beyond the per-token price.
FullsendAI
- On usage
- None
- Card top-up
- None
- Crypto top-up
- None
DeepInfra
- On usage
- None stated
- Card top-up
- None stated
- Crypto top-up
- No crypto option listed
- Source
- Source ·
OpenRouter
- On usage
- None stated
- Card top-up
- 5.5%, $0.80 minimum
- Crypto top-up
- 5%
- Source
- Source ·
| Provider | On usage | Card top-up | Crypto top-up | Source |
|---|---|---|---|---|
| FullsendAI | None | None | None | Our pricing |
| DeepInfra | None stated | None stated | No crypto option listed | Source · No platform or top-up fee stated. |
| OpenRouter | None stated | 5.5%, $0.80 minimum | 5% | Source · Fees are charged when you buy credits, not per request. |
Prices, model for model
USD per million tokens, input / output. DeepInfra's prices are after any promotion it was running; for OpenRouter, the price most of its providers charge, with the lowest-priced endpoint and its precision beneath.
DeepSeek V4.1 FlashDeepSeek
GLM 5.3 FlashZ.ai
GPT OSS 120BOpenAI
GPT OSS 20BOpenAI
Kimi K2.6Moonshot AI
Qwen3.8 MaxAlibaba
DeepSeek V4 Pro 0813DeepSeek
Claude Opus 5.5Anthropic
Gemini 3.1 ProGoogle
| Model | FullsendAI | DeepInfra | OpenRouter |
|---|---|---|---|
| DeepSeek V4.1 FlashDeepSeek | Not offered yet | $0.14 / $0.42Promotional pricedeepseek-ai/DeepSeek-V4.1-FlashSource · | $0.30 / $1.20from $0.022 / $0.192 (fp8, Morph)deepseek/deepseek-v4.1-flashSource · |
| GLM 5.3 FlashZ.ai | Not offered yet | $0.075 / $0.25Promotional pricezai-org/GLM-5.3-FlashSource · | $0.15 / $0.50from $0.0189 / $0.0625 (fp4, OpenInference)z-ai/glm-5.3-flashSource · |
| GPT OSS 120BOpenAI | Not offered yet | $0.037 / $0.17openai/gpt-oss-120bSource · | $0.15 / $0.60from $0.03 / $0.15 (fp4, Near AI)openai/gpt-oss-120bSource · |
| GPT OSS 20BOpenAI | Not offered yet | $0.03 / $0.14openai/gpt-oss-20bSource · | $0.07 / $0.15from $0.018 / $0.09 (fp8, Darkbloom)openai/gpt-oss-20bSource · |
| Kimi K2.6Moonshot AI | Not offered yet | $0.75 / $3.50moonshotai/Kimi-K2.6Source · | $0.95 / $4.00from $0.39 / $1.80 (int4, Chutes)moonshotai/kimi-k2.6Source · |
| Qwen3.8 MaxAlibaba | Not offered yet | $1.65 / $4.951Qwen/Qwen3.8-MaxSource · | $2.00 / $6.00qwen/qwen3.8-max-0902Source · |
| DeepSeek V4 Pro 0813DeepSeek | Not offered yet | $1.30 / $2.60deepseek-ai/DeepSeek-V4-Pro-0813Source · | $1.32 / $3.96from $0.675 / $1.20 (fp4, Sail Research)deepseek/deepseek-v4-pro-0813Source · |
| Claude Opus 5.5Anthropic | Not offered yet | $4.00 / $20.00anthropic/claude-opus-5-5Source · | $4.00 / $20.00anthropic/claude-opus-5.5Source · |
| Gemini 3.1 ProGoogle | Not offered yet | $2.00 / $12.00google/gemini-3.1-proSource · | $2.00 / $12.00google/gemini-3.1-pro-previewSource · |
What a month costs, fees included
Same model, same tokens, no cache hits, one card top-up for exactly what the tokens cost.
No model has a current price on every side today, so there is no like-for-like example to show. See the price table above.
Key differences
Host or router
DeepInfra runs the models itself, so you get one price and one place your requests run. OpenRouter forwards each request to one of many hosts and bills you in one place; it can fail over when a host is down, and the price depends on which host serves you.
Fees
DeepInfra states no platform fee. OpenRouter adds no markup to inference prices and charges 5.5% by card ($0.80 minimum) and 5% in crypto when you buy credits.
Precision
On OpenRouter the lowest-priced endpoints for many open models are quantized to fp4 or int4. DeepInfra lists a precision for each model it serves. Check the precision before comparing on price alone.
Migrate in 2 minutes
Change the base URL, the key and the model name. Everything else stays: streaming, tool calls and images work as before.
from openai import OpenAI
client = OpenAI(
base_url="https://fullsendai.com/v1",
api_key="fs_live_...",
)
reply = client.chat.completions.create(
model="deepseek-v4.1-flash",
messages=[{"role": "user", "content": "Hello"}],
)Full setup for every tool is in the quickstart and agent guides.
Questions
Is DeepInfra cheaper than OpenRouter?
Often, on the same open model: DeepInfra's own price is at or below the price most OpenRouter providers charge for most models in the table, and OpenRouter adds 5.5% by card ($0.80 minimum) and 5% in crypto when you buy credits. OpenRouter can route to cheaper, usually quantized, endpoints. Prices checked Oct 30, 2026.
Does OpenRouter use DeepInfra?
DeepInfra is one of the providers OpenRouter lists for many open models, in its public endpoints API. Which provider serves a given request depends on OpenRouter's routing and your settings.
Where does FullsendAI fit?
It's a third option for people who want more than an API: one key and balance for the API, a web chat and your coding agents, no top-up fee, crypto payments and every model in the USA.
Where does FullsendAI run models?
In the United States, every model. Requests are never routed abroad, not even when a host is busy.
Does FullsendAI keep my prompts?
Not from the API unless you ask: by default only what billing needs (tokens, cost, timing, status). If you turn on body storage for a project, its requests and replies are kept for 30 days so you can read them in Logs, then deleted.
Related comparisons
One key. Every model. No top-up fee.
Works with the OpenAI and Anthropic SDKs, Claude Code, Codex and other coding agents. Every model runs in the USA.