Comparison · DeepSeek vs Alibaba
DeepSeek V4 Pro vs Qwen3.8 Max: pricing, allowances and benchmarks
DeepSeek V4 Pro (DeepSeek) and Qwen3.8 Max (Alibaba): 2 shared subscriptions and 6 shared leaderboard snapshots, with workloads, sample dates and benchmark configurations.
Plan & benchmark data snapshot · How real prices are computed
As of 2026-10-16, 2 of 2 shared subscriptions use matching normalized workloads: DeepSeek V4 Pro has a lower recorded token cost on 2, Qwen3.8 Max on 0, with 0 tied. Their highest archived configurations score 36 and 40.15 on Artificial Analysis Intelligence Index v4.3; all 6 shared leaderboard snapshots, configurations and sources appear below.
- Shared subscriptions
- 2Both models available
- Comparable workloads
- 2Used for price ratios
- Shared leaderboards
- 6Highest archived configurations
- Official API prices
- 2/ 2Models with list prices
Plot both on the price–capability chart
Same plan, side by side
Each row is a subscription that includes both models at the same fee. Real cost = monthly fee ÷ the recorded monthly allowance. Each model's allowance assumes using the plan for that model; the columns cannot be added together. Click a price for its calculation and sources. Ratios use only matching normalized workloads.
| Plan / provider | Monthly fee | DeepSeek V4 Pro | Qwen3.8 Max | Recorded token cost |
|---|---|---|---|---|
| Command Code GOATCommand Code | $10/mo | $0.02387 /MTok 0.419B tokens/month Standard workload 2026-09-30 · official source date · confidence Medium | $0.1613 /MTok 0.062B tokens/month Standard workload 2026-09-30 · official source date · confidence Medium | DeepSeek V4 Pro lower6.76× price ratio (higher/lower) |
| OpenCode GoOpenCode | $10/mo | $0.03183 /MTok 0.314B tokens/month Standard workload 2026-09-30 · official source date · confidence Medium | $0.2151 /MTok 0.047B tokens/month Standard workload 2026-09-30 · official source date · confidence Medium | DeepSeek V4 Pro lower6.76× price ratio (higher/lower) |
Leaderboards head to head
Each model's highest archived configuration on each shared leaderboard snapshot, ranked among the models archived in this dataset. Harness, reasoning effort and service mode can differ; self-reported and estimated results are marked. These reference configurations are not measurements of the subscriptions above. Overlapping leaderboards do not form an overall ability rating.
| Leaderboard | Metric | DeepSeek V4 Pro | Qwen3.8 Max | Higher archived score |
|---|---|---|---|---|
| Artificial Analysis Intelligence Index v4.3 | Intelligence Index | 36 #32 / 60 · DeepSeek V4 Pro 0813 (max) Max 2026-10-01 · Record | 40.15 #22 / 60 · Qwen3.8 Max Max 2026-10-01 · Record | Qwen3.8 Max+4.16 |
| Terminal-Bench 4.0 (AA) | Resolution Rate % | 14.65 #24 / 53 · Artificial Analysis - DeepSeek V4 Pro (max) Artificial Analysis · Max 2026-10-01 · Record | 18.69 #23 / 53 · Artificial Analysis - Qwen3.8 Max Artificial Analysis · Max 2026-10-01 · Record | Qwen3.8 Max+4.04 |
| Code Arena | WebDev Overall | Arena Score | 1,582 #25 / 54 · deepseek-v4-pro-high-20260813 High Reported interval 1,572.7–1,591.3 2026-10-01 · Record | 1,671 #8 / 54 · qwen3.8-max Max Reported interval 1,659–1,683 2026-10-01 · Record | Qwen3.8 Max+89 |
| Agent Arena · Overall (Ranking / Pareto) | Net Improvement % | 1.29 #21 / 39 · DeepSeek V4 Pro (High) (0813) High Reported interval -0.41–2.99 2026-10-01 · Record | 2.8 #18 / 39 · Qwen3.8 Max Max Reported interval 2.15–3.45 2026-10-01 · Record | Qwen3.8 Max+1.51 |
| Artificial Analysis Coding Agent Index v1.5 | Coding Agent Index | 43.05 #16 / 19 · Codex - DeepSeek V4 Pro 0813 (max) Codex · Max 2026-10-01 · Record | 43.27 #14 / 19 · Claude Code - Qwen3.8 Max Claude Code · Max 2026-10-01 · Record | Qwen3.8 Max+0.21 |
| DeepSWE v1.1 | Pass@1 % | 62.83 #18 / 33 · mini-swe-agent - deepseek-v4-pro (max) mini-swe-agent · Max Reported interval 56.5–69.17 2026-09-12 · Record | 57.46 #20 / 33 · mini-swe-agent - qwen3-8-max (xhigh) mini-swe-agent · Max Reported interval 54.8–60.12 2026-09-12 · Record | DeepSeek V4 Pro+5.37 |
Official API pricing
Each row uses its listed pricing conditions. Cache writes are shown separately; subscription costs above retain their recorded quota basis. An em dash means that rate is not recorded.
| Model / pricing conditions | Cached input | Input | Cache write | Output |
|---|---|---|---|---|
| DeepSeek V4 ProUpstream referenceUpstream price reference | $0.044 | $1.32 | — | $3.96 |
| Qwen3.8 MaxStandard · international · implicit cache · Singapore · 1–1,000,000 input tokensOfficial source · Checked All pricing tiers | $0.25 | $2 | $2 | $6 |
How to read this comparison
Price ratios compare recorded USD/MTok under the same known workload. Raw measured totals, different workloads and unmetered access remain visible but are excluded from the lower-cost counts. Recorded allowances assume using the quota; your effective cost depends on utilization, cache mix, output mix and plan limits. Qualitative confidence is not an error bar.
- Standard workload: 97% cache reads / 2.5% input / 0.5% output