Dola-Seed 2.0 Mini, also referred to as Doubao-Seed-2.0-Mini internally, is the smallest and cheapest model in ByteDance's second-generation Seed 2.0 foundation model series. Where its siblings focus on frontier reasoning (Pro) and balanced production workloads (Lite), Mini was purpose-built for one thing: throughput at volume, with a price that removes the hesitation of scale.
Seed 2.0 Mini is what you reach for when you need to run a million documents through a model, process content at the edge, or build a product where AI is deeply embedded in every interaction. It's not the most capable model on the leaderboard. It's the one that makes the invoice feel manageable.
ByteDance doesn't sell a single model, it ships four distinct variants built on the same foundation, each carved out for a specific tier of use:
Seed 2.0 Mini is a heavily distilled version of the Pro model. The distillation process narrows the architecture, removing attention heads and reducing layer count, specifically to minimize inference latency and operating cost. Here are the key specs as of the February 2026 release:
| Specification | Value | Notes |
|---|---|---|
| Model ID | seed-2-0-mini-260215 | Date suffix = release version |
| Context Window | 32,000 tokens | Optimized for short-to-medium tasks |
| Multimodal Support | Text + Image | Inherited from Seed 2.0 backbone |
| Architecture | Distilled from Seed 2.0 Pro | Fewer attention heads, reduced layers |
| Tool Calling | Supported | Compatible with agent frameworks |
| API Compatibility | OpenAI-compatible | Minimal migration effort |
| Deployment | Cloud-only | Via Volcano Engine / BytePlus ModelArk |
| Release Date | February 14, 2026 | First in series live on BytePlus |
Mini's 32K window is smaller than Pro's 256K or Lite's 256K — a deliberate trade-off. For batch classification, content moderation, summarization, or any task involving documents that fit within 32K tokens, this is more than sufficient. Tasks requiring very long context should use Lite or Pro instead.
Is Dola-Seed 2.0 Mini the same as Doubao-Seed-2.0-Mini?
Yes. "Dola" is the international branding for ByteDance's AI products, while "Doubao" is the domestic Chinese branding. They refer to the same underlying models from the same Seed research team. "Dola-Seed 2.0 Mini" and "Doubao-Seed-2.0-Mini" are the same model — the naming depends on which market or platform you're accessing it through.
How does Seed 2.0 Mini compare to GPT-4o Mini?
Seed 2.0 Mini is positioned at a significantly lower price point than GPT-4o Mini, roughly 2× cheaper on input and output based on published rates as of April 2026. In terms of capability, both models are designed for high-throughput, cost-sensitive tasks rather than frontier reasoning. Seed 2.0 Mini has multimodal support and tool-calling, but GPT-4o Mini benefits from broader ecosystem tooling, wider developer familiarity, and strong English-language performance. For teams already on the OpenAI ecosystem, the switching cost is low, but validate on your specific workload first.
Does Seed 2.0 Mini support video input?
The broader Seed 2.0 family supports text, image, and video inputs — but Mini's multimodal capabilities are focused primarily on text and image understanding. For video analysis tasks, the Pro model with its leading VideoMME scores is the appropriate choice.
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
Dola-Seed 2.0 Mini is documented as suited for high-concurrency, batch-generation workloads where speed and deployment density matter. It is also suited for latency-sensitive, cost-sensitive, and lightweight tasks that prioritize fast response over peak capability.
It is documented for multimodal understanding, including accepting text, images, and video as input. It is also described as suitable for video analysis and document processing in high-volume API pipelines.
import requests r = requests.post( "https://api.aimlapi.com/v1/chat/completions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "bytedance/seed-2-0-mini", "messages": [ { "role": "user", "content": "Hello!" } ] }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/chat/completions", { method: "POST", headers: { Authorization: `Bearer ${process.env.AIMLAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "bytedance/seed-2-0-mini", "messages": [ { "role": "user", "content": "Hello!" } ] }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/chat/completions \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"bytedance/seed-2-0-mini","messages":[{"role":"user","content":"Hello!"}]}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
| Cached input |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| AIME 2025 | 87.0 | Competition mathematics (AIME), 2025 | Source | September 21, 2026 |
| GPQA Diamond | 79.0 | Google-proof graduate science questions (hardest subset) | Source | September 21, 2026 |
| MMLU-Pro | 83.6 | Multi-discipline knowledge + reasoning (harder MMLU) | Source | September 21, 2026 |
| LiveCodeBench | 64.1 | Contamination-free competitive programming problems | Source | September 21, 2026 |
| SWE-bench Verified | 67.9 | Resolving verified real GitHub issues | Source | September 21, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
Dola-Seed 2.0 Mini This page | Reasoning + agents | |||
| Reasoning + agents | ||||
| Balanced coding + agents | ||||
| Long-context, multimodal & agentic workflows | ||||
| Reasoning + agents |
Dola-Seed 2.0 Mini has a 256,000 tokens context window and can return up to 128,000 tokens.
Dola-Seed 2.0 Mini takes image, text, video as input and returns text.
Use bytedance/seed-2-0-mini as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
Dola-Seed 2.0 Mini is priced at input $0.13 / 1M tokens, output $0.52 / 1M tokens, cached input $0.026 / 1M tokens.
Yes, Dola-Seed 2.0 Mini can stream responses as they are generated.
Yes, Dola-Seed 2.0 Mini accepts image input alongside text.
Dola-Seed 2.0 Mini was built by ByteDance.
Yes. It supports function calling, parallel tool calls, and the Tools feature.
Send requests to https://api.aimlapi.com/v1/chat/completions using the model id bytedance/seed-2-0-mini.
