Alternatives to GPT-5.2-Codex

These models are alternatives by measurable coding evidence: they share the current closed-model classificationand are closest to GPT-5.2-Codex's SWE-bench score. Similarity here does not mean identical capabilities, licence terms, or output quality.

The target model

SWE-bench
72.4%
Cheapest input
$1.75
Cheapest output
$14.00
Hosts
3
AlternativeSWE-benchInput $/MTokOutput $/MTokHostsContextTradeoff
GLM 572.1%$0.60$1.9231203klower input price
GPT-4o (2024-08-06)71.9%$2.50$10.004128khigher input price
GPT-5.4 Mini73.0%$0.75$4.504400klower input price
Grok 4.371.4%$1.25$2.5021.0Mlower input price
MiniMax M2.773.8%$0.24$0.9621197klower input price
GLM 5.174.2%$0.91$2.8644203klower input price
MiniMax M2.574.2%$0.15$0.5931197klower input price
MiniMax M2.174.8%$0.29$0.958197klower input price
Claude Sonnet 4.570.0%$3.00$15.00101.0Mhigher input price
Gemini 3.5 Flash-Lite75.0%$0.15$1.254lower input price
MiniMax M375.0%$0.23$0.9628262klower input price
gpt-5.4-nano69.8%$0.20$1.254lower input price

How alternatives are selected

Alternatives must have a benchmark score and the same current open/closed classification as the target. They are sorted by score distance, then by price/performance. The table compares hosted API economics; it does not establish licence compatibility or self-hosting rights.

Check the model licence, provider Terms, context limits, tool support, and output quality before switching a production workload.