Baidu: ERNIE-Light-DT-1.0 vs Qwen3 8 Max — API Pricing, 150K Cache & Latency Benchmark
Side-by-side unit economics audit comparing Baidu: ERNIE-Light-DT-1.0 (baidu/ernie-light-dt-1.0-intl: Input $1.13/1M, Output $3.94/1M, Warm Cache Read $0.113/1M) against Qwen3 8 Max (qwen3-8-max: Input $0.8/1M, Output $2.4/1M, Warm Cache Read $0.08/1M).
Activate 4-breakpoint prefix caching on Baidu: ERNIE-Light-DT-1.0 to cut input token costs from $1.13/1M down to $0.113/1M, or route async workloads via Batch API for 50% off.
Lock static system prompt prefixes on Qwen3 8 Max to pay $0.08/1M warm cache rates instead of $0.8/1M uncached input, plus eligible startup cloud credits.
In Etuki FinOps telemetry comparing Baidu: ERNIE-Light-DT-1.0 (API ID: baidu/ernie-light-dt-1.0-intl) vs Qwen3 8 Max (API ID: qwen3-8-max), Qwen3 8 Max offers lower standard uncached input pricing at $0.8/1M tokens, while Qwen3 8 Max leads in 10-turn 150K cache-pinned coding economics at $0.3060 per 10 turns (saving $0.0948 over Baidu: ERNIE-Light-DT-1.0). For latency-sensitive interactive agents, Qwen3 8 Max delivers 112ms faster P90 time-to-first-token (320ms vs 432ms).
| Specification / FinOps Dimension | Baidu: ERNIE-Light-DT-1.0 (baidu/ernie-light-dt-1.0-intl) | Qwen3 8 Max (qwen3-8-max) | Verified Advantage |
|---|---|---|---|
| Official Vendor API ID | baidu/ernie-light-dt-1.0-intl | qwen3-8-max | UMCE 563-Registry Locked |
| 563-Registry & Routing Eligibility | is_known: true | routable: true | is_known: true | routable: true | Both Merkle Verified |
| Standard Uncached Input (/1M tokens) | $1.13 / 1M | $0.8 / 1M | Qwen3 8 Max |
| Warm Prefix Cache Read (/1M tokens) | $0.113 / 1M | $0.08 / 1M | Qwen3 8 Max |
| Standard Output (/1M tokens) | $3.94 / 1M | $2.4 / 1M | Qwen3 8 Max |
| Context Window Capacity | 1024K Tokens | 200K Tokens | Baidu: ERNIE-Light-DT-1.0 |
| 150K Prompt × 10 Turns (Unpinned Cold + 2K Out) | $1.7738 | $1.2480 | Qwen3 8 Max |
| 150K × 10 Turns Case A (100% Static Pinned) | $0.4008 (saves $1.3730) | $0.3060 (saves $0.9420) | Qwen3 8 Max (-$0.0948) |
| 150K × 10 Turns Case B (90% Static + 10% Tail) | $0.5381 (saves $1.2357) | $0.4032 (saves $0.8448) | Qwen3 8 Max |
| P50 / P90 First-Token Latency | 294ms / 432ms (89 tok/s) | 218ms / 320ms (105 tok/s) | Qwen3 8 Max (-112ms) |
1. Full-Turn Rate-Card Substitution for Baidu: ERNIE-Light-DT-1.0: At an uncached input rate of $1.13/1M, output rate of $3.94/1M, and warm cache read rate of $0.113/1M, running 10 consecutive 150,000-token coding agent turns (with 2,000 output tokens/turn) without prefix pinning costs $1.7738. Locking 100% static prefixes (Case A: 1 cold turn + 9 warm cached turns + 20K total output) reduces 10-turn spend to $0.4008 (saving $1.3730); under a 90% static prefix + 10% dynamic tail workload (Case B), 10-turn spend is $0.5381.
2. Full-Turn Rate-Card Substitution for Qwen3 8 Max & Winner Verdict: On Qwen3 8 Max ($0.8/1M input, $2.4/1M output, $0.08/1M warm cache), 10 unpinned 150K turns cost $1.2480, dropping to $0.3060 in Case A and $0.4032 in Case B. Across every 10-turn 150K coding loop (including output tokens), Qwen3 8 Max delivers an additional $0.0948 Case-A cost advantage.
Production Workload TCO: 25,000 MAU • 50M Tokens/Mo (50% Cache Hit) • 2,000 GB Egress
💡 Optimal FinOps Verdict: Deploying Qwen3 8 Max saves $22.66/month ($272/year) at 25,000 MAU & 50M monthly tokens (80% input / 20% output split with 50% prefix cache hit rate).