What other models does ByteDance offer?

ByteDance's Seed 2.0 family spans cost-optimized throughput to frontier-level reasoning, with a specialized coding variant and a GUI automation model rounding out the lineup. Here are some other models ByteDance supports:

• Dola Seed 2.0 Mini: ByteDance's entry-level model in the Seed 2.0 series, built for latency-sensitive, high-concurrency workloads. It supports a 256K context window at $0.10 input / $0.40 output per million tokens (as of 06/08/2026), and scores 87.0 on AIME 2025. It's best suited for classification, code completion, and structured dialogue at scale

• Dola Seed 2.0 Pro: ByteDance's flagship model for complex reasoning and agentic workflows. It scores 98.3 on AIME 2025, 88.9 on GPQA Diamond, and 3020 on Codeforces (as of 06/08/2026), with a 272K context window. It's designed for tasks that push the frontier of mathematical reasoning, video understanding, and multi-step planning

• Dola Seed 2.0 Code (preview): A code-specialized fine-tune of the Seed 2.0 base, with a 256K context window. It achieves 87.8 on LiveCodeBench v6 and 76.5% on SWE-Bench Verified (as of 06/08/2026) at $0.30 input / $1.20 output per million tokens, with optimizations like fill-in-the-middle completions and multi-file refactoring support

• UI-TARS-1.5-7B: ByteDance's 7B-parameter vision-language model built exclusively for GUI interaction. With a 128K context window and $0.10 input / $0.20 output pricing (as of 06/08/2026), it achieves 84.8% on WebVoyager and 46.6 on AndroidWorld. It's used for browser agents, desktop automation, and mobile interface navigation rather than general text generation

How does Dola Seed 2.0 Lite differ from ByteDance's other models?

Dola Seed 2.0 Lite occupies the middle tier of the Seed 2.0 lineup, delivering benchmark performance close to Pro while costing roughly half as much on input tokens.

• Pricing: At $0.25 per million input tokens and $2.00 per million output tokens (as of 06/08/2026), Lite costs approximately 2.5x less on input than Pro ($0.47 / $2.37) and about 2.5x more on input than Mini ($0.10 / $0.40). It's the balance point for teams that can't justify Pro pricing for every request but need more capability than Mini

• Context window: Lite supports 256K tokens (with tiered pricing above 128K), matching Mini and nearly matching Pro's 272K. This makes it viable for long-document summarization, extended code review sessions, and multi-turn agentic workflows without context truncation

• Benchmark scores: Lite scores 93.0 on AIME 2025, 85.1 on GPQA Diamond, 87.7 on MMLU-Pro, 2233 on Codeforces, and 73.5% on SWE-Bench Verified (as of 06/08/2026). On MMLU-Pro, Lite (87.7) actually edges out Pro (87.0), suggesting strong general knowledge and instruction-following for everyday tasks

• Use case fit: Lite supports multimodal inputs including text, image, and video, plus tool-calling for agent pipelines. It's engineered for high-frequency visual understanding and agentic workflows, not edge deployments or cost-at-all-costs scenarios.

Lite is the right choice for production applications that need reliable multimodal reasoning and agentic capability across a high volume of requests, where Pro-tier pricing would make unit economics unworkable.

What models should I consider using alongside Dola Seed 2.0 Lite?

No single model is optimal for every task. Here are models worth pairing with Dola Seed 2.0 Lite depending on what your product needs:

• Dola Seed 2.0 Pro When requests involve frontier-level mathematical reasoning (AIME-class problems), complex multi-step planning, or video understanding tasks where the gap between Lite and Pro matters, route to Pro. Lite handles the bulk of production traffic while Pro absorbs the reasoning-heavy tail

• Dola Seed 2.0 Code (preview) For dedicated software engineering tasks, code agents, or IDE integrations where fill-in-the-middle and multi-file refactoring optimizations matter, routing to Code over Lite saves roughly 58% on input costs while maintaining equivalent SWE-Bench performance

• Claude Sonnet 4.5 When tasks require particularly strong instruction adherence on complex structured output schemas or nuanced long-form writing, Claude Sonnet 4.5 is a cross-provider complement. It also gives you provider redundancy if ByteDance's platform has availability issues

• Gemini 2.5 Flash For tasks where you need a 1M-token context window, such as processing full codebases or book-length documents in a single pass, Gemini 2.5 Flash extends beyond Lite's 256K ceiling at competitive pricing. Use it as a routing target when context length is the binding constraint

• Mistral Small For European regulatory requirements (GDPR, data residency), Mistral Small provides a low-cost alternative that keeps data within EU infrastructure. Pairing it with Lite gives you geographic routing flexibility without sacrificing too much on quality for standard tasks

What are the challenges of using Dola Seed 2.0 Lite in my product?

Like any production LLM, Dola Seed 2.0 Lite comes with tradeoffs worth planning for:

• Output pricing at long contexts: Lite's output cost rises from $2.00 to $4.00 per million tokens for requests exceeding 128K input tokens (as of 06/08/2026). For long-document workflows that generate substantial output, this tiered pricing structure can materially increase per-request costs compared to expectations based on the base rate

• Coding depth vs. Code variant: While Lite scores 73.5% on SWE-Bench Verified (as of 06/08/2026), it lacks the fill-in-the-middle optimizations and coding-specific kernel tuning in Dola Seed 2.0 Code. For software engineering agents that need clean inline completions and minimal explanatory interruptions, Lite may produce noisier outputs

• Provider dependency: Concentrating requests on a single ByteDance endpoint creates fragility if the platform experiences downtime, rate limiting, or deprecates this model version. Without a configured fallback, your application has no recovery path during an outage

• Cost at scale: At $2.00 per million output tokens, Lite is affordable for moderate volumes but compounds quickly at enterprise scale. Without hard budget caps and per-project attribution, output costs can exceed projections as request volume grows

• Knowledge cutoff: Lite's training data has a January 2024 cutoff. Applications that depend on current events, recent library documentation, or post-cutoff technical releases need retrieval augmentation or real-time grounding to avoid stale or incorrect responses

Why should I use Merge Gateway to route LLM requests with Dola Seed 2.0 Lite and every other model?

Using Dola Seed 2.0 Lite through Merge Gateway gives you access to the model itself and the infrastructure layer around it:

• Intelligent routing and automatic failover: Merge routes around ByteDance outages automatically. Routing policies based on cost, latency, or quality can reduce spend by 40–60% without touching your application code. Lite's tiered pricing above 128K context makes cost-aware routing especially valuable here

• One API, every provider: Access Dola Seed 2.0 Lite and every other major LLM through a single endpoint and API key. Change providers by swapping the model string—no application code changes required

• Cost governance: Set hard or soft project budgets so Dola Seed 2.0 Lite spend stays within plan. Every request is attributed to a model, project, and tag in a unified billing dashboard across all providers

• Build Your Own Router: Define what "best" means for your traffic by selecting from curated ML benchmarks or adding your own eval scores. The router scores each available model against your weights and picks the winner per request, with a plain-language explanation of every decision

• Security and compliance controls: Apply DLP rules and prompt injection protection before every request reaches ByteDance. Enforce per-project model and region policies without adding that logic to your application

How can I start routing requests to Dola Seed 2.0 Lite via Merge Gateway?

Getting Dola Seed 2.0 Lite running through Merge Gateway takes a few minutes:

1. Create an account and get your API key from the dashboard.

2. Install the Merge Gateway SDK: run pip install merge-gateway-sdk (Python) or npm install merge-gateway-sdk (Node). Alternatively, if you're already using the OpenAI SDK, set base_url = "https://api-gateway.merge.dev/v1/openai" and your existing code works as-is.

3. Make your first request using the provider/model format. For Dola Seed 2.0 Lite, the model string is tencent/dola-seed-2.0-lite. Swap the model string to route to any other provider without changing anything else.

4. Configure a routing policy in the dashboard to set failover behavior, cost limits, and optimization strategy. Your first policy can be as simple as naming Dola Seed 2.0 Lite as primary with one fallback.

Full setup instructions and SDK references are in the Merge Gateway docs.