Seed 2.0 Code

Seed 2.0 Code sits inside ByteDance's Doubao/Seed family and is positioned as a developer-facing assistant tuned for everyday coding environments. Third-party catalogs describe it as a coding-first model that aims for reliable tool use inside common IDE shells, with particular strength on front-end generation tasks where visual polish and structured markup matter. Its design emphasis is less on open-ended chat and more on acting as a pragmatic coding teammate that can read a workspace, call out edits, and follow multi-step development workflows end to end.

The model is served through several gateway-style providers, including DeepInfra and Volcano Engine (Volcengine), the latter being ByteDance's own distribution channel for Doubao-family models, which suggests closer alignment with the original training pipeline. It is documented with multimodal input handling, reasoning, tool calling, structured outputs, and caching, giving it the breadth needed to plug into agentic coding setups that combine file inspection, code edits, and verification steps. Pricing varies meaningfully across endpoints, with per-million-token input costs ranging roughly from the mid-forties to the low-fifties of a dollar, so teams running heavy IDE sessions can choose the provider whose latency and cost profile best matches their development loop.

Merge Gatewaybytedance/dola-seed-2.0-codeseed

Open provider page Provider docs

Quick Info

Provider: Merge Gateway
Model key: bytedance/dola-seed-2.0-code
Release date: Feb 14, 2026
Last updated: Feb 14, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost: $0.40
Output token cost: $2.40

Limits

Output tokens: 128,000 tokens
Context window: 256,000 tokens

Transparent token rates

Compare Seed 2.0 Code pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Seed 2.0 Code

Merge Gateway

CoverageBenchmark

According to llm-stats.com, Seed 2.0 Code was released on April 29, 2026 by ByteDance under a proprietary license, and is listed as multimodal, accepting both text and images as input. The tracked provider surface for direct hosted access is DeepInfra, where pricing starts at $0.50 per million input tokens, $0.10 per m The product page also lists related ByteDance variants for comparison, including Seed 2.1 Pro, Seed 2.0 Pro, Seed 2.0 Mini, Seed 2.0 Lite, and Seed 1.8, treating these as distinct models with their own pages. Notably, llm-stats.com names the tracked provider as DeepInfra, not Merge Gateway, so this candidate provides t

Merge Gateway

CoverageBenchmark

naga.ac operates a dedicated Seed-2.0-Code Chat Completions performance page that explicitly names the exact model variant "Seed-2.0-Code" (model id: seed-2.0-code) and attributes it to ByteDance Seed. The page, created August 12, 2026, documents a 262.1k token context window for the Chat Completions endpoint and chara The page surfaces throughput data through early September 2026 and indicates that time-to-first-token (TTFT) telemetry was not yet sufficient to publish, while offering links to a Playground, an Examples section, an API Reference, and an Activity/Performance dashboard. Because naga.ac is a third-party inference host ra

Merge Gateway

Coverage

The Baidu encyclopedia entry confirms that the Doubao-Seed-2.0 series, comprising Pro, Lite, Mini, and a Code model, was launched on February 14, 2026, and is optimized for large-scale production environment requirements with efficient reasoning, multimodal understanding, and complex instruction execution. Pro is desig The Code variant is specifically tailored for programming scenarios, integrated into the TRAE China Edition, and is described as having superior reasoning cost compared with most comparable models. The series is distributed as API services via Volcano Engine, with the Pro model already deployed across the Doubao app on

Merge Gateway

CoverageRelease Notes

The EmpirioLabs changelog, relevant because EmpirioLabs is the listed host for Seed 2.0 Code on Merge Gateway, documents several technical changes through 2026. The August 26, 2026 entry announces that EmpirioLabs AI is now listed as a provider on Opper, a European AI gateway based in Stockholm, with Opper managing the The July 7, 2026 changelog entry ships two infrastructural features that affect how developers interact with EmpirioLabs-hosted models like Seed 2.0 Code: a Batch API offering 35% off list price via /v1/files and /v1/batches endpoints (each line targeting /v1/chat/completions or /v1/embeddings), and a Gemini-compatible