tokevenMid-tier Models
6 balanced models compared across Claude, GPT, Gemini, and Grok, sorted by input price.
Mid-tier models balance quality and cost for most production workloads at scale. They handle everyday tasks well and cost significantly less than flagship options.
Mid-tier Models
By task
Best mid-tier model per task
FAQ
What are mid-tier models?
Mid-tier Models are the balanced performance-and-cost AI models. Currently 6 models in this tier across Claude, GPT, Gemini, and Grok.
Which mid-tier model is cheapest?
Grok 4.3 is the cheapest at $1.25/$2.50 per million tokens (input/output).
Confidence scores are a third-party benchmark snapshot from week 29 of 2026. This snapshot is 8 weeks old, so treat the scores as directional and validate against your own evals.
Price every mid-tier model
Per-call cost calculators for each model in this tier.
Grok 4.3 cost calculator
$1.25 / $2.50 per MTok
Claude Sonnet 5 cost calculator
$2.00 / $10.00 per MTok
GPT 5.6 Terra cost calculator
$2.00 / $12.00 per MTok
Gemini 3.1 Pro cost calculator
$2.00 / $12.00 per MTok
GPT 5.4 cost calculator
$2.50 / $15.00 per MTok
Claude Sonnet 4.6 cost calculator
$3.00 / $15.00 per MTok
Other tiers and comparisons
Track your real costs
Comparisons show estimated costs at list price. The dashboard reveals your real spend and savings.