Yi 1.5 34B
01.AI · dense · 34.4B parameters · 200,000 context
Parameters
34.4B
Context Window
195K tokens
Architecture
Dense
Best GPU
H20
Cheapest API
$0.80/M
Quality Score
72/100
Intelligence Brief
Yi 1.5 34B is a 34.4B parameter DENSE model from 01.AI, featuring Grouped Query Attention (GQA) with 60 layers and 7,168 hidden dimensions. With a 200,000 token context window, it supports tools, structured output, code, math, multilingual. On standardized benchmarks, it achieves MMLU 76.8, HumanEval 44, GSM8K 78. The most cost-effective API deployment is via 01ai at $0.80/M output tokens. For self-hosted inference, H20 delivers optimal throughput at $940/month.
Provider pricing
1 provider · canonical: 01ai| Provider | Input $/M | Output $/M ▲ | Notes |
|---|---|---|---|
| 01aicanonical | $0.800 | $0.800 | cheapest input · cheapest output |
Prices update via the nightly pricing cron + admin approvals at /admin/ingest-queue. The leaderboard's Input/Output cells show the canonical rate above; this table shows the full spread.
Recent changes
Loading…
Related models
5 suggestions- Yi 1.5 9BYi 1.5 · 8.83B$0.200/M out
- Yi-LargeYi · 24B$3.00/M out
- Yi-LightningYi · 22B$0.990/M out
Qwen 2.5 14BQwen 2.5 · 14.8B$0.300/M out- Qwen 2.5 32BQwen 2.5 · 32.5B$0.800/M out
Picks: same family first, then same vendor within ±2× params, then top tag-overlap matches. Price shown is the cheapest Output $/M across providers — the row's page shows the canonical anchor.
Architecture Details
Memory Requirements
BF16 Weights
68.8 GB
FP8 Weights
34.4 GB
INT4 Weights
17.2 GB
GPU Compatibility Matrix
Yi 1.5 34B is compatible with 57% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
FP8 · 1 GPU · tensorrt-llm
100/100
score
Throughput
972.8 tok/s
Latency (ITL)
1.0ms
Est. TTFT
0ms
Cost/Month
$940
Cost/M Tokens
$0.37
FP8 · 1 GPU · tensorrt-llm
98/100
score
Throughput
1.1K tok/s
Latency (ITL)
1.0ms
Est. TTFT
0ms
Cost/Month
$4261
Cost/M Tokens
$1.54
FP8 · 1 GPU · tensorrt-llm
95/100
score
Throughput
1.1K tok/s
Latency (ITL)
1.0ms
Est. TTFT
0ms
Cost/Month
$2553
Cost/M Tokens
$0.93
Deployment Options
API Deployment
01ai
$0.80/M
output tokens
Single GPU
H20
$940/mo
Min VRAM: 34 GB
Multi-GPU
RTX A6000 x2
106.4 tok/s
TP· $930/mo
API Pricing Comparison
| Provider | Input $/M | Output $/M | Badges |
|---|---|---|---|
| 01ai | $0.80 | $0.80 | Cheapest |
Cost Analysis
| Provider | Input $/M | Output $/M | ~Monthly Cost |
|---|---|---|---|
| 01aiBest Value | $0.80 | $0.80 | $8 |
Cost per 1,000 Requests
Short (500 tok)
$0.56
via 01ai
Medium (2K tok)
$2.24
via 01ai
Long (8K tok)
$8.00
via 01ai
Performance Estimates
Throughput by GPU
VRAM Breakdown (H20, FP8)
Precision Impact
bf16
68.8 GB
weights/GPU
fp8
34.4 GB
weights/GPU
~972.8 tok/s
int4
17.2 GB
weights/GPU
Quality Benchmarks
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy Yi 1.5 34B
Self-Hosted Infrastructure
Similar Models
Qwen-AgentWorld-35B-A3B
34.7B params · moe
Quality: 50
Code Llama 34B
34B params · dense
Quality: 55
from $0.78/M
Aya 23 35B
35B params · dense
Quality: 50
from $1.50/M
Command R (August 2024)
35B params · dense
Quality: 68
from $0.60/M
Command R
35B params · dense
Quality: 68
from $0.50/M
Frequently Asked Questions
How much VRAM does Yi 1.5 34B need for inference?
Yi 1.5 34B requires approximately 68.8 GB of VRAM at BF16 precision, 34.4 GB at FP8, or 17.2 GB at INT4 quantization. Additional VRAM is needed for KV-cache (245760 bytes per token) and activations (~2.00 GB).
What is the best GPU for Yi 1.5 34B?
The top recommended GPU for Yi 1.5 34B is the H20 using FP8 precision. It achieves approximately 972.8 tokens/sec at an estimated cost of $940/month ($0.37/M tokens). Score: 100/100.
How much does Yi 1.5 34B inference cost?
Yi 1.5 34B API inference starts from $0.80/M input tokens and $0.80/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.