Loading…
Route text, image, video and music through a single OpenAI-compatible endpoint. Claude, GPT, Gemini, Grok, Qwen, Seedance, Kling and Veo — with automatic failover, spend controls and no KYC. Up to 80% below list price.
Free credit on signup · prepaid in USDT · cancel by simply not topping up
Live on the gateway
0
Models live
0
API formats, one key
0
Video models live
0
KYC required
Powering production traffic for builders, agencies and automation teams
The newest frontier models, live on the gateway.
GPT-6 Luna
Newgpt-6-luna
Gemini 3.8 Flash
Newgemini-3.8-flash
Grok 4.20
Newgrok-4.20
GPT-6.1 Sol
Newgpt-6.1-sol
GPT-6 Astra
Newgpt-6-astra
GPT-5.5
Newgpt-5.5
GPT-5.4 Pro
Newgpt-5.4-pro
Claude Opus 5.5
Newclaude-opus-5-5
Every frontier family, one key
150 models live on the gateway across text, image, video and audio. These are the model families we serve — this page makes no claim of partnership or endorsement, and every one of them is reachable with the same key.
Generate video and stills from the same key: Seedance, Kling, Wan, Veo and the rest of the 150-model catalog. POST /v1/videos, poll, done.
Change one line — base_url and api_key. Your existing OpenAI, Anthropic, or LangChain code keeps working.
Talk OpenAI (/v1/chat/completions) or Anthropic (/v1/messages) with the same key. The Anthropic Messages surface is served natively by HELIOROUTE.
Top up with USDT on Binance Smart Chain (BEP-20). No card, no paperwork, no waiting for approval.
Compare
Same models you already use — minus the procurement, the KYC and the per-vendor invoices.
| HELIOROUTE | Buying direct | Other routers | |
|---|---|---|---|
| One key for text, image and video | – | partial | |
| OpenAI + Anthropic formats natively | partial | partial | |
| Automatic failover across vendors | – | ||
| Prepaid, no contract | |||
| Signup without KYC or card | – | partial | |
| Pay in USDT | – | partial | |
| 27+ video models on tap | – | – | |
| Sub-keys and reseller console | – | partial |
For teams
Volume pricing
Committed spend tiers down to cost + 15%. Rates locked for the term.
Dedicated capacity
Reserved routing and priority queues — your traffic never waits behind retail.
Invoiced billing
Monthly invoice, net terms, PO support. No prepaid balance required at scale.
Named support
Direct channel to the engineers on call. Response SLA written into the agreement.
99.9%
Uptime target on committed plans
4
Independent upstreams, auto-failover
$0
Platform fee — you pay for usage only
No sales call, no contract, no KYC. Sign up and start calling — text or video.
Email only. You get free credit immediately — enough to test chat and video before you pay anything.
Create account →One key works for OpenAI (/v1/chat/completions), Anthropic (/v1/messages) and video (/v1/videos).
Read the docs →Change base_url and api_key. Every SDK, LangChain, LlamaIndex and Cursor keeps working unchanged.
Top up with USDT →Catalog
Text, reasoning, coding, vision, image, video and audio — 150 models, all on the same key, all metered per request.
150 models
PVT-003
pvt-003
$1.20/1M
−0%
Hermes 3 405B
unc-hermes-3-405b
$0.040/1M
−0%
Hermes 4 405B
unc-hermes-4-405b
$0.040/1M
−80%
GPT-6 Luna
gpt-6-luna
$0.020/1M
−80%
Gemini 3.8 Flash
gemini-3.8-flash
$0.150/1M
−80%
Grok 4.20
grok-4.20
$0.250/1M
−80%
Grok 4.7
grok-4.7
$0.400/1M
−80%
GPT-6.1 Sol
gpt-6.1-sol
$0.400/1M
−80%
GPT-6 Astra
gpt-6-astra
$2.00/1M
−80%
GPT-6 Sol
gpt-6-sol
$0.400/1M
−80%
GPT-5.5
gpt-5.5
$1.00/1M
−80%
GPT-5.4 Pro
gpt-5.4-pro
$6.00/1M
−80%
Claude Opus 5.5
claude-opus-5-5
$0.800/1M
−80%
Claude Fable 5.1
claude-fable-5-1
$2.00/1M
−80%
Grok 4.20 Multi-Agent
grok-4.20-multi-agent
$0.250/1M
−80%
Claude Sonnet 5.5
claude-sonnet-5-5
$0.400/1M
−80%
GLM 5.3 PrimeOSS
glm-5.3-prime
$0.112/1M
−80%
GLM-5.2OSS
glm-5.2
$0.280/1M
−80%
DeepSeek V4.1 FlashOSS
deepseek-v4-1-flash
$0.060/1M
−80%
DeepSeek V4 ProOSS
deepseek-v4-pro
$0.264/1M
−80%
Qwen3.8 Max PrimeOSS
qwen3.8-max-prime
$0.160/1M
−80%
Kimi K3OSS
kimi-k3
$0.600/1M
−80%
Qwen3 MaxOSS
qwen3-max
$0.240/1M
−80%
MiniMax M3OSS
MiniMax-M3
$0.060/1M
−80%
GPT Image 2.5 Flare
ws-ima-openai-gpt-image-25-flare
$0.007/img
−0%
GPT Image 2.5 Sunburst
ws-ima-openai-gpt-image-25-sunburst
$0.007/img
−0%
Nano Banana 2
kie-img-nano-banana-2
$0.003/img
−0%
Nano Banana Pro
kie-img-nano-banana-pro
$0.005/img
−0%
Nano Banana
kie-img-nano-banana
$0.002/img
−0%
Grok Imagine
ws-ima-xai-grok-imagine-image
$0.007/img
−0%
Recraft 20B
ws-ima-recraftai-recraft-20b
$0.007/img
−0%
Veo 3.1
kie-vid-veo-3-1
$0.030/clip
−0%
Video
Generate clips from the models you already pay for elsewhere. POST a prompt, poll the id, get an MP4 back — one bill, one endpoint, OpenAI-compatible request shape.
“luxury perfume bottle on black marble, water droplets, macro product commercial”
“drone shot over misty mountain peaks at sunrise, golden volumetric light”
“glowing blue cube rotating on dark reflective surface, volumetric light, slow motion”
“abstract blue light trails forming a network, premium tech commercial”
Veo 3.1
kie-vid-veo-3-1
Veo 3
ws-veo-3
Seedance 2.5
ws-seedance-2.5
Kling 2.5
ws-kling-2.5
Grok Imagine Video
kie-vid-grok-imagine
Wan 3.0
ws-wan-3.0
Seedance 1.0 Pro Fast
ws-vid-bytedance-seedance-v1-pro-fast
LTX 2 Pro
ws-vid-lightrick-ltx-2-pro
Audio
Speech and music run on the same key, the same balance and the same request style as everything else — priced per track, so there is no token maths and no per-second surprise. Music renders as a finished track, not a short preview.
audio.session
elevenlabs-tts
warm narration, british male
elevenlabs-tts
product walkthrough, calm pace
elevenlabs-tts
on-hold message, friendly tone
Uncensored
Alongside the frontier models we serve open-weight instruction models with their safety layers removed — Hermes, Dolphin, WizardLM and Mythomax. They are research and fiction models: they will follow the brief instead of returning a refusal, which makes them the right tool for security work, red-teaming, adversarial testing and unrestricted creative writing. Same key, same balance, same request format as everything else.
uncensored.family
unc-hermes-3-405b
Hermes 3 405B
unc-hermes-4-405b
Hermes 4 405B
unc-hermes-3-70b
Hermes 3 70B
unc-wizardlm-2-8x22b
WizardLM 2 8x22B
unc-dolphin-mistral-24b
Dolphin Mistral 24B
Real prices on both sides. No hidden fees, no unit games.
200K context
per month
Claude Sonnet 5: $0.400 in · $2.000 out per 1M tokens
Official: $2.000 in · $10.000 out per 1M tokens
Official
$40.00
HELIOROUTE
$8.00
You save 80% versus buying direct.
$32.00 saved per month at this volume.
Suggested top-up
Starter — $10.00 for $12.00 credit (+20%)
Pay-as-you-go. Bigger top-ups carry a bigger bonus.
Point your existing client at our base URL and use your HELIOROUTE key. That is the whole migration.
Base URL: https://helioroute.com/v1
sign in at https://helioroute.com
curl https://helioroute.com/v1/chat/completions \
-H "Authorization: Bearer $HELIOROUTE_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-sonnet-5",
"messages": [{"role":"user","content":"hello"}]
}'from openai import OpenAI
client = OpenAI(
base_url="https://helioroute.com/v1",
api_key="sk-hr-...",
)
resp = client.chat.completions.create(
model="claude-sonnet-5",
messages=[{"role": "user", "content": "hello"}],
)
print(resp.choices[0].message.content)Create an account (email), copy your API key from the console, and point your OpenAI client at our base URL. Your account starts with a small free credit — no card needed.
Yes. The API is OpenAI-compatible: set base_url to our /v1 endpoint and use your HELIOROUTE key. The Anthropic Messages format is served from the same host, so Claude SDK clients work too.
You buy credit up front. Every request is metered per token by the gateway and deducted from your balance in real time. The console shows spend per model, per day, and per key — and every chat reply shows its own token count and cost.
Requests stop with a clear budget-exceeded error. Your keys stay intact — top up any time and traffic resumes immediately. No surprise overage bills.
Only models we have verified end-to-end are listed as available. Anything unverified is shown as "coming soon", never as live.
No prompt content is logged by the panel. We store request metadata (model, token counts, cost, status) for billing and the usage dashboard.
Crypto only: USDT on Binance Smart Chain (BEP-20), settled by our own on-chain watcher. No UPI, no cards, no KYC.
Credits are prepaid. Unused credit is not refundable except where we agree in writing or where the law requires it. Contact support from the email you signed up with.
One key for text, image, video and music. Free credit on signup, prepaid in USDT, no KYC, cancel by simply not topping up.
Questions? support@helioroute.com