Claude Max subscription
$200.00/mo
List price used as the subsidy anchor.
Market distortion warning
The $200.00 Max plan is treated as a subsidized baseline. Focus on pressure ratio: (cost vs Max plan) / (intelligence vs Opus). Lower is more dangerous. High pressure now requires near-Opus intelligence and API cost that is close to Max-plan economics.
Critical: 1 model is subsidy breaker risk
MiniMax M2.5 deliver near- Opus intelligence at a tiny fraction of Opus-equivalent cost.
True API value
$4000.00
Opus-equivalent range target: $3000.00-$5000.00.
Subsidy gap
$3800.00
About 20.0x the plan price in Opus API value.
Usage volume configuration
Token values are in millions per month and set the Opus reference spend. Flagship rows use fixed market-chart monthly values; custom rows use live token math.
Current Opus-equivalent baseline is $4000.00 (within the $3,000-$5,000 subsidized range).
Model comparison matrix
Opus baseline is pinned first; all other rows are sorted by pressure value (highest threat first).
| Model | I/O $/M | Monthly | Intel / SWE | vs Opus | Pressure | Status | |
|---|---|---|---|---|---|---|---|
Claude Opus 4.6Anthropic | $5.00 / $25.00 | $4000.00 $39.60/intel | 101.0 81.5% SWE | 100.0% 100.0% | - | Baseline | |
MiniMax M2.5MiniMax | $0.30 / $1.20 | $240.00 $2.35/intel | 102.0 80.2% SWE | 101.0% 6.0% | 1.19x score 91.9 | Breaker | |
Qwen 3.5Alibaba | $0.25 / $1.00 | $190.00 $1.97/intel | 96.5 75.8% SWE | 95.5% 4.8% | 0.99x score 55.4 | Severe | |
GLM-5Zhipu AI | $1.00 / $3.20 | $720.00 $7.35/intel | 98.0 77.8% SWE | 97.0% 18.0% | 3.71x score 20.8 | Low | |
Codex 5.3OpenAI | $1.75 / $14.00 | $1400.00 $13.59/intel | 103.0 82.8% SWE | 102.0% 35.0% | 6.86x score 7.1 | Low | |
Gemini 3.1 ProGoogle | $2.00 / $12.00 | $1650.00 $16.50/intel | 100.0 81% SWE | 99.0% 41.3% | 8.33x score 3.3 | Low | |
GPT-5.2OpenAI | $1.75 / $14.00 | $1400.00 $14.58/intel | 96.0 73% SWE | 95.0% 35.0% | 7.36x score 2.9 | Low | |
Claude Sonnet 4.6Anthropic | $3.00 / $15.00 | $2400.00 $24.00/intel | 100.0 79.6% SWE | 99.0% 60.0% | 12.12x score 0.6 | Low | |
Grok 4.2xAI | $3.00 / $15.00 | $2400.00 $24.49/intel | 98.0 72.5% SWE | 97.0% 60.0% | 12.37x score 0.4 | Low |
Status legend
BreakerPressure value 70+ (highest threat).
SeverePressure value 45-69.9 (serious threat).
EmergingPressure value 25-44.9 (watch closely).
LowEither too expensive for the intelligence delivered or still too far from Opus.
BaselineOpus reference anchor.
Methodology notes
- Subsidized baseline target is Opus-equivalent spend in the $3,000-$5,000 range against the Max plan list price.
- Pressure ratio = (model monthly cost / Max plan cost) / (model intelligence / Opus intelligence).
- Lower pressure ratio is stronger: 1.0x means near-Max-plan cost at Opus-level intelligence.
- Flagship rows use Feb 2026 reference monthly cost values from the provided chart.
- Pressure value uses an intelligence gate (must exceed 90% of Opus) multiplied by an exponential cost penalty from pressure ratio.
- Table is sorted by pressure value descending, with the Opus baseline pinned to the first row.
- SWE-Bench remains visible as the direct coding benchmark.
- Intelligence score = weighted average (SWE 25%, Long Horizon 15%, AIME 20%, GPQA 15%, LiveCodeBench 25%).
- Cost per intelligence point = monthly cost / intelligence score.