Haiku vs GPT-4o-mini vs Gemini Flash (cost tier)

The cheap-and-fast models from each major lab.

For triage, classification, and high-volume transforms, the cost-tier models from each lab — Claude Haiku 4.5, GPT-4o mini, Gemini 2.5 Flash — all clear the quality bar most teams need at a fraction of flagship cost. Pick by latency profile and tool-use behaviour.

OpenAI: GPT-4o-mini
openai
Context
128K
Max output
16,384
Input
$0.15/Mtok
Output
$0.60/Mtok
Modalities
text, image, file
Gemini 2.5 Flash
google
Context
1049K
Max output
65,536
Input
$0.15/Mtok
Output
$0.60/Mtok
Modalities
text, image, video, audio

Other showdowns