Gemini 2.x
LLM Provider / Model · Huge documents, video/audio understanding, long-context retrieval
At a glance
| Input price | $1.25 |
| Output price | $5 |
| Cache price | $0.31 |
| Price tier | Mid |
| Context window | 1M+ |
| Max output | 8K |
| Context tier | 500K+ |
| Speed tier | Fast |
| Latency | fast |
| Knowledge cutoff | Aug 2024 |
| Modality | Multimodal (vision + audio) |
| Model ID | gemini-2.0-pro |
| Provider | |
| HumanEval | 86% |
| MMLU | 85% |
| SWE-Bench | 52% |
| Benchmark | mid-high |
| Released | Dec 2024 |
| Hosting | Closed/API |
| Capabilities | Vision, Audio, Tool use, Streaming, Structured output, Long context |
What Gemini 2.x does
Vision, Audio, Tool use, Streaming, Structured output, Long context
Best for
Huge documents, video/audio understanding, long-context retrieval
Works well with
Integration
Agent / Orchestration
Conflicts & caveats
- Kiro's spec mode is tuned for Claude models — Gemini 2.x may work but spec-driven flows are best with Claude.
- OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family — Gemini 2.x is not selectable inside this integration.
- Claude Code only works with Claude models — your LLM is Gemini 2.x.
- Devin uses Cognition's proprietary model internally — your selected Gemini 2.x will be ignored.
- OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected Gemini 2.x will be ignored.
- The Claude Code SDK is tuned for Claude models — Gemini 2.x may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
- Lovable uses a fixed internal model — your selected Gemini 2.x will be ignored.
- Replit Agent 3 uses its own model selection internally — your selected Gemini 2.x will be ignored.
Build a full stack around Gemini 2.x — Flowpicker shows compatibility warnings before you commit.
Open the stack planner →