One agent session
1M input tokens and 100K output tokens, no caching.
- Input1M x $0.25 = $0.25
- Output0.1M x $2 = $0.20
$0.45per session
ByteDance Seed - Closed weights
Seed 2.0 Lite pricing and specs
Seed 2.0 Lite is a closed-weights model from ByteDance Seed that costs $0.25 per 1M input tokens and $2 per 1M output tokens on the first-party API, with a 256K-token context window.
Last verified from docs.byteplus.com. All sources
- Input / 1M
- $0.25
- Output / 1M
- $2
- Context
- 256K tokens
- Weights
- Closed
Pricing per 1M tokens
| Item | Price / 1M tokens |
|---|---|
| InputPrompts up to 128,000 tokens | $0.25 |
| OutputPrompts up to 128,000 tokens | $2 |
| Cache readPrompts up to 128,000 tokens | $0.05 |
| Long-context inputPrompts over 128,000 tokens | $0.50 |
| Long-context outputPrompts over 128,000 tokens | $4 |
| Long-context cache readPrompts over 128,000 tokens | $0.10 |
Note: Current 260428 version. Prices are for non-audio input on BytePlus ModelArk online inference (standard), tiered by prompt length: [0, 128K] at base pricing, (128K, 256K] at long-context pricing. Audio input is billed separately: 3.75 input / 0.75 cache-hit up to 128K, 7.50 / 1.50 above. ModelArk bills cache storage at 0.0083 USD per M tokens per hour instead of a cache write fee, so cache write is not stated. The batch inference table does not list this version, so batch prices are blank. The model list states context length 256K, max input 256K and max output 128K including chain of thought (default 4K). BytePlus writes K as 1,000. Release notes describe it as the first omnimodal understanding model in the Dola family (video, image, audio and text). The deprecation notice names it dola-seed-2-0-lite-260428 and makes it the replacement for seed-1-8, seed-1-6 and seed-2-0-lite-260228. The release date is left blank because the release notes list it only under month 202605, with no day. No Seed 2.0 id appears in the Vercel AI Gateway model list.
Specs
- Context window
- 256,000 tokens
- Max output
- 128,000 tokens
- Provider
- ByteDance Seed
- API model id
seed-2-0-lite-260428- Weights
- Closed weights
What Seed 2.0 Lite costs in practice
Worked from the standard list rates above, with the math shown so you can swap in your own token counts. These use the base rates, which apply while each request's prompt stays at or under 128,000 tokens; above that, Seed 2.0 Lite bills $0.50 input and $4 output for the whole request.
A chat app
10,000 requests a day at 2K input and 500 output tokens each.
- Input20M x $0.25 = $5.00
- Output5M x $2 = $10.00
$15.00per day
$450.00 per 30-day month
The same session at a 90% cache hit
90% of the input is read from the prompt cache. The one-time cache write premium is not included.
- Cached input0.9M x $0.05 = $0.045
- Uncached input0.1M x $0.25 = $0.025
- Output0.1M x $2 = $0.20
$0.27per session
Compare with
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Seed 2.0 Lite (this page) | $0.25 | $2 |
| GPT-3.5 Turbo InstructOpenAI | $1.50 | $2 |
| GPT-5 MiniOpenAI | $0.25 | $2 |
| Grok Build 0.1xAI | $1 | $2 |
| Qwen3.5-35B-A3BAlibaba Cloud (Qwen) | $0.25 | $2 |
| Dola Seed 2.1 TurboByteDance Seed | $0.50 | $2.50 |
| Seed 2.0 ProByteDance Seed | $0.50 | $3 |
| Seed 2.0 MiniByteDance Seed | $0.10 | $0.40 |
FAQ
How much does Seed 2.0 Lite cost per million tokens?
Seed 2.0 Lite costs $0.25 per 1M input tokens and $2 per 1M output tokens on ByteDance Seed's first-party API, per docs.byteplus.com on October 17, 2026. Per 1M tokens, cache reads cost $0.05. Prompts over 128,000 tokens are billed at $0.50 input and $4 output for the whole request.
What is Seed 2.0 Lite's context window?
Seed 2.0 Lite has a 256,000-token context window and returns up to 128,000 output tokens per response.
Are Seed 2.0 Lite's weights open?
No. Seed 2.0 Lite is a closed-weights model. You call it through ByteDance Seed's API as seed-2-0-lite-260428.
Sources
Last verified from docs.byteplus.com.
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.