Alternatives to Claude Opus 4.6
These models are alternatives by measurable coding evidence: they share the current closed-model classificationand are closest to Claude Opus 4.6's SWE-bench score. Similarity here does not mean identical capabilities, licence terms, or output quality.
The target model
SWE-bench
78.7%
Cheapest input
$5.00
Cheapest output
$25.00
Hosts
1
| Alternative | SWE-bench | Input $/MTok | Output $/MTok | Hosts | Context | Tradeoff |
|---|---|---|---|---|---|---|
| GLM-5.2 | 78.7% | $1.20 | $4.20 | 4 | — | lower input price |
| GPT-5.6 Terra | 78.4% | $2.00 | $12.00 | 1 | 1.1M | lower input price |
| Kimi-K2.6 | 76.7% | $0.73 | $3.50 | 3 | 262k | lower input price |
| GPT-5.6 Luna | 75.7% | $0.20 | $1.20 | 1 | 1.1M | lower input price |
| Gemini 3 Flash Preview | 75.4% | $0.50 | $3.00 | 1 | — | lower input price |
| Claude Sonnet 4.6 | 75.2% | $3.00 | $15.00 | 1 | 1.0M | lower input price |
| GPT-5.5 | 83.1% | $5.00 | $30.00 | 1 | 1.1M | closest measured match |
| Claude Opus 4.7 | 83.5% | $5.00 | $25.00 | 1 | 1.0M | closest measured match |
| GLM 5 | 72.1% | $1.20 | $3.84 | 6 | — | cheaper, lower coding score |
| Gemini 2.5 Pro | 57.6% | $1.25 | $10.00 | 1 | — | cheaper, lower coding score |
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.