Skip to content

Same-model provider benchmark

DeepInfra vs Phala: LLM provider comparison

Compare DeepInfra and Phala on 16 exact shared text models. ProviderBench keeps speed, price, and catalog coverage separate so naturally faster model catalogs cannot distort the result.

DeepInfra

87 indexed models

Headquarters
United States
Server regions
N/a
Model types
67 text, 15 embeddings, 5 speech

Phala

18 indexed models

Headquarters
United States
Server regions
N/a
Model types
18 text

At a glance

Metric winners

There is no overall score. Each winner answers one specific question using only directly comparable data.

Fastest on shared models

1.51× typical advantage

16 exact models with complete recent speed data

Lowest token cost

$0.8318 / 1M

16 exact models using a 1K-input/500-output mix

Most models available

87 models

Complete catalog coverage across all indexed modalities

Shared-model benchmark summary

MetricDeepInfraPhala
Typical 500-token response19.11 s22.47 s
Typical response start0.78 s2.00 s
Typical output speed28.5 tok/s25.0 tok/s
Blended token price$0.8318 / 1M$1.1494 / 1M
Typical route uptime97.95%98.19%

Visual comparison

Price and performance charts

DeepInfraPhala
500-token response by shared model

Estimated seconds using recent median response-start and output-speed data. Lower is better.

DeepInfra has the lower typical same-model response ratio across 16 measured models.

Blended token price by shared model

USD per 1 million tokens using a 1,000-input/500-output mix. Lower is better.

DeepInfra has the lower typical price ratio across 16 priced shared models.

Shared text models

16 exact models · newest first

ModelDeepInfraPhala
GLM 5.2Winner · Phala

z-ai/glm-5.2

DeepInfra
Blended price
$1.62 / 1M
500-token response
42.37 s
Response starts
0.70 s
Output speed
12.0 tok/s
Route uptime
N/a
Context
1M
Route
deepinfra/fp4
Phala
Blended price
$2.4 / 1M
500-token response
16.81 s
Response starts
2.10 s
Output speed
34.0 tok/s
Route uptime
86.82%
Context
1M
Route
phala
Qwen3.6 27BWinner · DeepInfra

qwen/qwen3.6-27b

DeepInfra
Blended price
$1.28 / 1M
500-token response
27.04 s
Response starts
0.73 s
Output speed
19.0 tok/s
Route uptime
99.57%
Context
262.1K
Route
deepinfra/fp8
Phala
Blended price
$1.1133 / 1M
500-token response
39.31 s
Response starts
0.85 s
Output speed
13.0 tok/s
Route uptime
99.71%
Context
262.1K
Route
phala
Kimi K2.6Winner · DeepInfra

moonshotai/kimi-k2.6

DeepInfra
Blended price
$1.6667 / 1M
500-token response
12.84 s
Response starts
0.93 s
Output speed
42.0 tok/s
Route uptime
99.94%
Context
262.1K
Route
deepinfra/fp4
Phala
Blended price
$2.26 / 1M
500-token response
27.84 s
Response starts
2.84 s
Output speed
20.0 tok/s
Route uptime
N/a
Context
262.1K
Route
phala
GLM 5.1Winner · DeepInfra

z-ai/glm-5.1

DeepInfra
Blended price
$1.8667 / 1M
500-token response
18.81 s
Response starts
1.57 s
Output speed
29.0 tok/s
Route uptime
99.92%
Context
202.8K
Route
deepinfra/fp4
Phala
Blended price
$2.2067 / 1M
500-token response
38.16 s
Response starts
22.53 s
Output speed
32.0 tok/s
Route uptime
N/a
Context
202.8K
Route
phala
Gemma 4 31BWinner · DeepInfra

google/gemma-4-31b-it

DeepInfra
Blended price
$0.2033 / 1M
500-token response
22.57 s
Response starts
0.83 s
Output speed
23.0 tok/s
Route uptime
99.06%
Context
262.1K
Route
deepinfra/turbo
Phala
Blended price
$0.2533 / 1M
500-token response
53.66 s
Response starts
3.66 s
Output speed
10.0 tok/s
Route uptime
52.78%
Context
262.1K
Route
phala
Qwen3.5-27BWinner · Phala

qwen/qwen3.5-27b

DeepInfra
Blended price
$1.04 / 1M
500-token response
19.41 s
Response starts
1.55 s
Output speed
28.0 tok/s
Route uptime
93.77%
Context
262.1K
Route
deepinfra/fp8
Phala
Blended price
$1 / 1M
500-token response
18.65 s
Response starts
1.99 s
Output speed
30.0 tok/s
Route uptime
97.83%
Context
262.1K
Route
phala
Qwen3.5 397B A17BWinner · DeepInfra

qwen/qwen3.5-397b-a17b

DeepInfra
Blended price
$1.3 / 1M
500-token response
14.53 s
Response starts
0.64 s
Output speed
36.0 tok/s
Route uptime
79.03%
Context
262.1K
Route
deepinfra/fp8
Phala
Blended price
$1.5333 / 1M
500-token response
28.75 s
Response starts
3.75 s
Output speed
20.0 tok/s
Route uptime
99.36%
Context
262.1K
Route
phala
MiniMax M2.5Winner · DeepInfra

minimax/minimax-m2.5

DeepInfra
Blended price
$0.4833 / 1M
500-token response
11.21 s
Response starts
1.01 s
Output speed
49.0 tok/s
Route uptime
100.00%
Context
196.6K
Route
deepinfra/fp8
Phala
Blended price
$0.5933 / 1M
500-token response
15.91 s
Response starts
2.02 s
Output speed
36.0 tok/s
Route uptime
N/a
Context
196.6K
Route
phala
GLM 5Winner · DeepInfra

z-ai/glm-5

DeepInfra
Blended price
$1.0933 / 1M
500-token response
21.01 s
Response starts
1.01 s
Output speed
25.0 tok/s
Route uptime
99.79%
Context
202.8K
Route
deepinfra/fp4
Phala
Blended price
$1.9667 / 1M
500-token response
22.96 s
Response starts
2.96 s
Output speed
25.0 tok/s
Route uptime
N/a
Context
202.8K
Route
phala
Kimi K2.5Winner · DeepInfra

moonshotai/kimi-k2.5

DeepInfra
Blended price
$1.05 / 1M
500-token response
16.56 s
Response starts
0.93 s
Output speed
32.0 tok/s
Route uptime
99.96%
Context
262.1K
Route
deepinfra/fp4
Phala
Blended price
$1.4 / 1M
500-token response
21.68 s
Response starts
1.68 s
Output speed
25.0 tok/s
Route uptime
89.90%
Context
262.1K
Route
phala
GLM 4.7Winner · DeepInfra

z-ai/glm-4.7

DeepInfra
Blended price
$0.85 / 1M
500-token response
13.65 s
Response starts
0.83 s
Output speed
39.0 tok/s
Route uptime
99.92%
Context
202.8K
Route
deepinfra/fp4
Phala
Blended price
$1.6667 / 1M
500-token response
21.70 s
Response starts
1.70 s
Output speed
25.0 tok/s
Route uptime
N/a
Context
131.1K
Route
phala
DeepSeek V3.2Winner · DeepInfra

deepseek/deepseek-v3.2

DeepInfra
Blended price
$0.3 / 1M
500-token response
42.39 s
Response starts
0.73 s
Output speed
12.0 tok/s
Route uptime
83.26%
Context
163.8K
Route
deepinfra/fp4
Phala
Blended price
$1 / 1M
500-token response
66.48 s
Response starts
3.98 s
Output speed
8.0 tok/s
Route uptime
98.54%
Context
163.8K
Route
phala

qwen/qwen3-vl-30b-a3b-instruct

DeepInfra
Blended price
$0.3 / 1M
500-token response
45.94 s
Response starts
0.48 s
Output speed
11.0 tok/s
Route uptime
96.83%
Context
262.1K
Route
deepinfra/fp8
Phala
Blended price
$0.3667 / 1M
500-token response
21.98 s
Response starts
1.15 s
Output speed
24.0 tok/s
Route uptime
90.96%
Context
128K
Route
phala
gpt-oss-120bWinner · DeepInfra

openai/gpt-oss-120b

DeepInfra
Blended price
$0.0813 / 1M
500-token response
12.10 s
Response starts
0.47 s
Output speed
43.0 tok/s
Route uptime
94.64%
Context
131.1K
Route
deepinfra/bf16
Phala
Blended price
$0.3 / 1M
500-token response
12.10 s
Response starts
0.74 s
Output speed
44.0 tok/s
Route uptime
99.76%
Context
131.1K
Route
phala
gpt-oss-20bWinner · DeepInfra

openai/gpt-oss-20b

DeepInfra
Blended price
$0.0667 / 1M
500-token response
5.15 s
Response starts
0.35 s
Output speed
104.0 tok/s
Route uptime
99.78%
Context
131.1K
Route
deepinfra/bf16
Phala
Blended price
$0.0767 / 1M
500-token response
9.39 s
Response starts
0.62 s
Output speed
57.0 tok/s
Route uptime
99.78%
Context
131.1K
Route
phala
Gemma 3 27BWinner · DeepInfra

google/gemma-3-27b-it

DeepInfra
Blended price
$0.1067 / 1M
500-token response
30.09 s
Response starts
0.68 s
Output speed
17.0 tok/s
Route uptime
99.95%
Context
131.1K
Route
deepinfra/fp8
Phala
Blended price
$0.2533 / 1M
500-token response
73.29 s
Response starts
1.86 s
Output speed
7.0 tok/s
Route uptime
44.04%
Context
262.1K
Route
phala

A per-model winner combines blended price and estimated 500-token response time with equal proportional weight. Ties and rows missing either measurement receive no badge. Comparison data calculated . Values use one deterministic route per provider and model; missing measurements remain visible as N/a.

DeepInfra vs Phala analysis

How DeepInfra and Phala compare for AI inference

DeepInfra and Phala share 16 indexed text models, including GLM 5.2, Qwen3.6 27B, Kimi K2.6, GLM 5.1, Gemma 4 31B. DeepInfra has the stronger typical response-time result on the directly measured set.

Same-model speed evidence

16 shared models currently have complete response-start and output-speed measurements on both providers. The speed comparison uses per-model ratios before taking the median, so naturally faster model catalogs do not improve the result.

Token pricing on one workload

16 shared models have complete input and output prices on both providers. Prices use the same 1,000-input/500-output-token mix and are normalized to one million tokens for readability.

Catalog and deployment differences

DeepInfra has 87 indexed models and lists no published server regions; Phala has 18 models and lists no regions. Verify data residency, privacy terms, limits, and production latency directly before choosing.