Deal of the week
DeepSeek V4 Flash — our deepest cut
$0.14 / $0.28 per 1M tokens is already the cheapest flagship-class pricing on the market. We take another 30% off.
See the price42+ models · 15 vendors · one endpoint
One API.Every frontier model.
TokDash routes OpenAI, Anthropic, Google, DeepSeek and 11 more vendors through a single OpenAI-compatible endpoint. Official list prices, a transparent per-model discount, and uptime you can check before you commit.
Prepaid credits · No subscription · OpenAI-compatible
curl https://api.tokdash.com/v1/chat/completions \
-H "Authorization: Bearer $TOKDASH_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"messages": [{"role": "user", "content": "Explain CRDTs in two sentences."}]
}'from openai import OpenAI
client = OpenAI(
api_key="YOUR_TOKDASH_KEY",
base_url="https://api.tokdash.com/v1",
)
resp = client.chat.completions.create(
model="claude-sonnet-5",
messages=[{"role": "user", "content": "Summarize this changelog."}],
)
print(resp.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.TOKDASH_API_KEY,
baseURL: "https://api.tokdash.com/v1",
});
const res = await client.chat.completions.create({
model: "gemini-3.5-flash",
messages: [{ role: "user", content: "Draft a release note." }],
});export ANTHROPIC_BASE_URL="https://api.tokdash.com"
export ANTHROPIC_AUTH_TOKEN="$TOKDASH_API_KEY"
claude42+Models
−24%Avg. discount
99.9%30-day uptime
15Vendors
1.28BTokens routed this month
Highlights
Day-0 launch
Sora 2 is live on day one
Listed the day it shipped. $0.10 per second of video, with the model-specific discount applied from your first request.
Generate a videoStudio offer
Volume pricing without the sales call
Commit $2,000 a month and every model drops to −25%. One invoice, one API key, no per-vendor contracts.
Compare plansModels & pricing
Every price below is the official list price next to ours. Recalculate any row — the discount is the only thing we change.
Chat & multimodal
5/26DeepSeek V4 Flashchat
$0.14in$0.28outper 1M tokens
$0.098 / $0.196−30%Save $0.084/1M output
99.95%548 ms median
DeepSeek V4chat
$0.55in$2.20outper 1M tokens
$0.396 / $1.58−28%Save $0.616/1M output
99.95%555 ms median
Claude Sonnet 5chat
$3.00in$15.00outper 1M tokens
$2.40 / $12.00−20%Save $3.00/1M output
99.95%560 ms median
Gemini 3.5 Flashchat
$1.50in$9.00outper 1M tokens
$1.20 / $7.20−20%Save $1.80/1M output
99.9%584 ms median
Gemini 3.5 Prochat
$2.50in$15.00outper 1M tokens
$2.05 / $12.30−18%Save $2.70/1M output
99.95%540 ms median
GPT-5.6 Minichat
$0.60in$3.00outper 1M tokens
$0.51 / $2.55−15%Save $0.45/1M output
99.9%490 ms median
GPT-5.6 Solchat
$5.00in$30.00outper 1M tokens
$4.25 / $25.50−15%Save $4.50/1M output
99.95%524 ms median
Grok 5chat
$3.00in$15.00outper 1M tokens
$2.55 / $12.75−15%Save $2.25/1M output
99.95%560 ms median
Claude Fable 5chat
$10.00in$50.00outper 1M tokens
$8.80 / $44.00−12%Save $6.00/1M output
99.95%536 ms median
GPT-5.6 o3chat
$2.00in$10.00outper 1M tokens
$1.76 / $8.80−12%Save $1.20/1M output
99.95%620 ms median
Qwen 3.5 Flashchat
$0.10in$0.40outper 1M tokens
$0.07 / $0.28−30%Save $0.12/1M output
99.85%490 ms median
GLM 5chat
$0.7353in$2.94outper 1M tokens
$0.5294 / $2.12−28%Save $0.8235/1M output
99.9%545 ms median
MiniMax M3chat
$0.5882in$2.35outper 1M tokens
$0.4235 / $1.69−28%Save $0.6588/1M output
99.85%535 ms median
Moonshot K2chat
$1.47in$5.88outper 1M tokens
$1.06 / $4.24−28%Save $1.65/1M output
99.9%550 ms median
Qwen 3.5 Maxchat
$0.2941in$1.76outper 1M tokens
$0.2118 / $1.27−28%Save $0.4941/1M output
99.9%560 ms median
Claude Haiku 5chat
$0.80in$4.00outper 1M tokens
$0.60 / $3.00−25%Save $1.00/1M output
99.9%500 ms median
DeepSeek V4 R1chat
$0.80in$3.20outper 1M tokens
$0.60 / $2.40−25%Save $0.80/1M output
99.95%580 ms median
Doubao Prochat
$0.2941in$1.18outper 1M tokens
$0.2206 / $0.8824−25%Save $0.2941/1M output
99.85%500 ms median
Gemma 5chat
$0.20in$1.00outper 1M tokens
$0.15 / $0.75−25%Save $0.25/1M output
99.85%510 ms median
Qwen 3.5 Turbochat
$0.35in$1.40outper 1M tokens
$0.2625 / $1.05−25%Save $0.35/1M output
99.9%520 ms median
Gemini 3.5 Nanochat
$0.30in$1.50outper 1M tokens
$0.234 / $1.17−22%Save $0.33/1M output
99.85%480 ms median
Llama 5chat
$0.80in$4.00outper 1M tokens
$0.624 / $3.12−22%Save $0.88/1M output
99.9%530 ms median
Claude Sonnet 5 Minichat
$0.80in$4.00outper 1M tokens
$0.64 / $3.20−20%Save $0.80/1M output
99.9%520 ms median
GPT-5.6 Nanochat
$0.15in$0.60outper 1M tokens
$0.12 / $0.48−20%Save $0.12/1M output
99.9%470 ms median
Grok 5 Minichat
$0.60in$3.00outper 1M tokens
$0.48 / $2.40−20%Save $0.60/1M output
99.9%510 ms median
Mistral Large 3chat
$2.00in$6.00outper 1M tokens
$1.60 / $4.80−20%Save $1.20/1M output
99.9%540 ms median
Image generation
5/6Nano Banana Proimage
$0.13per image
$0.1066−18%Save $0.0234/image
99.85%7 s median
Seedream 4.0image
$0.0294per image
$0.0206−30%Save $0.0088/image
99.85%7 s median
FLUX 1.1 Proimage
$0.04per image
$0.03−25%Save $0.01/image
99.85%7 s median
Nano Banana 2image
$0.045per image
$0.0351−22%Save $0.0099/image
99.85%7 s median
GPT Image 1image
$0.04per image
$0.032−20%Save $0.008/image
99.85%7 s median
Grok Image 2image
$0.05per image
$0.04−20%Save $0.01/image
99.8%7 s median
Video generation
5/6Sora 2video
$0.10per second
$0.085−15%Save $0.015/sec
99.55%55 s median
Hailuo 2video
$0.15per second
$0.1125−25%Save $0.0375/sec
99.5%55 s median
Kling 3.0video
$0.10per second
$0.075−25%Save $0.025/sec
99.55%55 s median
Veo 3.1 Fastvideo
$0.15per second
$0.12−20%Save $0.03/sec
99.55%55 s median
Veo 3.1 Standardvideo
$0.40per second
$0.34−15%Save $0.06/sec
99.55%55 s median
Sora 2 Provideo
$0.30per second
$0.264−12%Save $0.036/sec
99.55%55 s median
Audio
4/4MiniMax Speech 2.6audio
$29.41per 1M chars
$21.18−28%Save $8.24/1M chars
99.85%900 ms median
GPT-4o mini TTSaudio
$12.00per 1M chars
$9.60−20%Save $2.40/1M chars
99.85%900 ms median
Whisper ASRaudio
$0.006per minute
$0.0048−20%Save $0.0012/min
99.85%900 ms median
ElevenLabs v3audio
$5.00per month
$4.25−15%Save $0.75/mo
99.85%900 ms median
No models match
Request a model →Built for teams that ship, not for script resellers
TokDash is operated like infrastructure: monitored endpoints, explicit pricing, one bill.
24/7 gateway operations
Every endpoint is probed continuously and watched by engineers around the clock. Most incidents are handled before they reach your users.
Private by default
Prompts and completions are not stored and not used for training. Traffic routes through Singapore and US-West. No Chinese account or ID required — for any model.
One invoice, every vendor
A single account covers all vendors, with one monthly invoice. Pay by card, wire transfer or USDT. Invoices and receipts are generated automatically.
Day-0 models, 24-hour price sync
New flagship models are listed the day they launch. When a provider changes its list price, this page reflects it within 24 hours — with the change noted.
Trusted by teams shipping with AI
- GDPR-ready
- Zero retention
- SOC 2 in progress
- 99.9% SLA
We moved three products onto one endpoint and cut our model spend by a third. The pricing page is the only one our finance team has ever approved without questions.
What would you save?
Pick a model and a monthly volume. The estimate uses a 3:1 input-to-output ratio by default.
Official cost$113
TokDash cost$96
You save$17$203 per year
Estimate only. Actual billing depends on your real input/output mix.
Change one line. Keep your code.
TokDash is a drop-in replacement for the OpenAI SDK. Point the base URL at us and use the same request shape you already ship.
curl https://api.tokdash.com/v1/chat/completions \
-H "Authorization: Bearer $TOKDASH_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"messages": [{"role": "user", "content": "Explain CRDTs in two sentences."}]
}'from openai import OpenAI
client = OpenAI(
api_key="YOUR_TOKDASH_KEY",
base_url="https://api.tokdash.com/v1",
)
resp = client.chat.completions.create(
model="claude-sonnet-5",
messages=[{"role": "user", "content": "Summarize this changelog."}],
)
print(resp.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.TOKDASH_API_KEY,
baseURL: "https://api.tokdash.com/v1",
});
const res = await client.chat.completions.create({
model: "gemini-3.5-flash",
messages: [{ role: "user", content: "Draft a release note." }],
});export ANTHROPIC_BASE_URL="https://api.tokdash.com"
export ANTHROPIC_AUTH_TOKEN="$TOKDASH_API_KEY"
claudeFrequently asked questions
Is TokDash an official reseller?
We are an independent API gateway. We buy capacity directly from providers and route your requests through our infrastructure. Official list prices are shown unchanged; our margin comes only from the discount we apply to each model, which is printed on this page.
Do you store my prompts or completions?
No. Requests are proxied in memory and are not written to disk or used for training. Operational logs keep latency, status codes and token counts — never message content.
How does billing work?
TokDash is prepaid. You top up your balance first, and each request deducts the discounted price from it — no subscription, no monthly minimum, no surprise invoice. When the balance runs out, requests stop instead of running up a bill. Top up by card through Stripe, by wire transfer for larger amounts, or with USDT.
How fast are new models added?
Flagship models are usually listed on launch day. Provider price changes are reflected here within 24 hours, and the "prices last synced" timestamp at the bottom of every page tells you when we last checked.
Can I keep using my existing OpenAI SDK?
Yes. Change the base URL to https://api.tokdash.com/v1 and the model name. Request and response shapes, streaming, tool calls and embeddings stay the same.
Stop paying list price for the same tokens.
Create a key, run one request, and compare your bill at the end of the month.