Baseten: Models Intelligence, Performance & Price

Baseten
Baseten

This analysis is intended to support you in choosing the best model provided by Baseten for your use-case.

Most Intelligent

#1
Kimi K3 (max)
Kimi K3 (max)
60
#2
DeepSeek V4 Pro 0813 (max)
DeepSeek V4 Pro 0813 (max)
53
#3
GLM-5.2 (max) (FAST)
GLM-5.2 (max) (FAST)
53
#4
GLM-5.2 (max)
GLM-5.2 (max)
53
#5
DeepSeek V4 Flash 0731 (max)
DeepSeek V4 Flash 0731 (max)
52

Intelligence index

Total 19 models

Fastest

#1
Inkling Small
Inkling Small
338 t/s
#2
GLM-5.2 (max) (FAST)
GLM-5.2 (max) (FAST)
286 t/s
#3
gpt-oss-120b (low)
gpt-oss-120b (low)
278 t/s
#4
gpt-oss-120b (high)
gpt-oss-120b (high)
276 t/s
#5
Inkling
Inkling
236 t/s

Output speed

Total 19 models

Lowest Price

#1
DeepSeek V4 Flash 0731 (max)
DeepSeek V4 Flash 0731 (max)
$0.07
#2
gpt-oss-120b (high)
gpt-oss-120b (high)
$0.14
#3
gpt-oss-120b (low)
gpt-oss-120b (low)
$0.14
#4
Inkling Small
Inkling Small
$0.29
#5
GLM-4.7
GLM-4.7
$0.42

Blended price (per 1M tokens)

Total 19 models

Indicates a reasoning model

Baseten offers 19 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.

  • For intelligence, the top models on Baseten are Kimi K3 (max) (60), DeepSeek V4 Pro 0813 (max) (53), and GLM-5.2 (max) (FAST) (53).
  • For output speed, the fastest models are Inkling Small (338 t/s), GLM-5.2 (max) (FAST) (286 t/s), and gpt-oss-120b (low) (278 t/s).
  • For latency, DeepSeek V4 Pro (0.60s), GLM-4.7 (0.65s), and GLM-5.2 (FAST) (1.83s) offer the lowest time to first answer token.
  • For pricing, DeepSeek V4 Flash 0731 (max) ($0.07), gpt-oss-120b (high) ($0.14), and gpt-oss-120b (low) ($0.14) offer the lowest blended prices per 1M tokens. Prices vary up to 5.9x across models.
  • For context window size, Kimi K3 (max) (1M), GLM-5.2 (max) (1M), and DeepSeek V4 Flash 0731 (max) (1M) support the largest context windows on Baseten.

Highlights

Artificial Analysis Intelligence Index · Higher is better
Output tokens per second · Higher is better
USD per 1M tokens (blended) · Lower is better

Intelligence Evaluations

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.1.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR
Reasoning models are indicated by a lightbulb icon

Artificial Analysis Intelligence Index v4.1.1 includes: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better
See more

Agentic real-world work tasks, (Elo-500)/2000

Agentic tool use

Agentic coding & terminal use

Coding

Reasoning & knowledge

Scientific reasoning

Physics reasoning

Long context reasoning

Agentic knowledge work, Elo

Agentic SaaS workflows

Legal agentic work, criterion pass rate

Agentic business operations

Quantitative analysis on spreadsheets & documents

Instruction following

Long-horizon agentic tasks

Kubernetes incident root-cause analysis

Visual reasoning

Reasoning models are indicated by a lightbulb icon

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

Artificial Analysis Intelligence Index v4.1.1 includes: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line
Reasoning models are indicated by a lightbulb icon

While higher intelligence models are typically more expensive, they do not all follow the same price-quality curve.

Context Window

Context Window

Context window: tokens limit · Higher is better
Reasoning models are indicated by a lightbulb icon

Larger context windows are relevant to RAG (Retrieval Augmented Generation) LLM workflows which typically involve reasoning and information retrieval of large amounts of data.

Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).

Pricing

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line
Reasoning models are indicated by a lightbulb icon

While higher intelligence models are typically more expensive, they do not all follow the same price-quality curve.

Performance Summary

Output Speed vs. Price

Output speed: output tokens per second · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line
Reasoning models are indicated by a lightbulb icon

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Speed

Measured by Output Speed (tokens per second)

Output Speed

Output tokens per second · Higher is better
Reasoning models are indicated by a lightbulb icon

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).

Latency

Measured by Time (seconds) to First Token

Latency: Time To First Answer Token

Seconds to first answer token received · Accounts for reasoning model 'thinking' time
Reasoning models are indicated by a lightbulb icon

Time to first answer token received, in seconds, after API request sent. For reasoning models, this includes the 'thinking' time of the model before providing an answer. For models which do not support streaming, this represents time to receive the completion.

End-to-End Response Time

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

End-to-End Response Time vs. Price

End-to-end response time: end-to-end seconds to output 500 tokens · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line
Reasoning models are indicated by a lightbulb icon

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Further Analysis
Kimi logo
Kimi K3 (max)
1.05M
Open
60
$0.84
101
0.88
25.52
19.72
DeepSeek logo
DeepSeek V4 Pro 0813 (max)
1M
Open
53
$1.12
182
1.01
14.78
11.02
Z AI logo
GLM-5.2 (max) (FAST)
524k
Open
53
$0.54
286
2.23
10.98
7.00
Z AI logo
GLM-5.2 (max)
1.05M
Open
53
$0.35
106
2.17
25.84
18.94
DeepSeek logo
DeepSeek V4 Flash 0731 (max)
1.05M
Open
52
$0.14
153
0.88
17.24
13.10
Kimi logo
Kimi K3 (low)
1M
Open
48
$0.34
76
0.76
33.48
26.18
DeepSeek logo
DeepSeek V4 Pro (max)
1M
Open
45
$0.44
159
0.65
31.22
27.44
Kimi logo
Kimi K2.6
262k
Open
45
$0.37
195
0.69
26.07
22.82
DeepSeek logo
DeepSeek V4 Pro (high)
1M
Open
44
$0.44
159
0.64
16.33
12.54
Kimi logo
Kimi K2.7 Code
262k
Open
43
$0.29
224
0.46
12.64
9.95
Thinking Machines logo
Inkling
262k
Open
42
$0.39
236
0.63
11.20
8.46
Thinking Machines logo
Inkling Small
262k
Open
41
$0.10
338
0.47
7.86
5.91
Kimi logo
Kimi K2.6
262k
Open
35*
--
--
--
--
--
Z AI logo
GLM-5.2 (FAST)
524k
Open
35
--
212
1.83
4.19
--
Z AI logo
GLM-5.2
1.05M
Open
35
--
80
2.33
8.62
--
Z AI logo
GLM-4.7
200k
Open
34
$0.14
204
0.65
12.92
9.81
DeepSeek logo
DeepSeek V4 Pro
1M
Open
32*
--
167
0.60
3.60
--
Z AI logo
GLM-4.7
200k
Open
27*
--
204
0.65
3.10
--
OpenAI logo
gpt-oss-120b (high)
128k
Open
24
$0.05
276
0.24
9.31
7.26
OpenAI logo
gpt-oss-120b (low)
128k
Open
15
$0.01
278
0.22
9.20
7.18

Key definitions

Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).

Frequently Asked Questions

Common questions about Baseten

The most intelligent model available on Baseten is Kimi K3 (max) with an Intelligence Index score of 60.

The fastest model on Baseten by output speed is Inkling Small at 338.4 tokens per second.

The model with the lowest time to first answer token on Baseten is DeepSeek V4 Pro at 0.60s. Lower latency means faster initial response time.

The most affordable model on Baseten by blended price is DeepSeek V4 Flash 0731 (max) at $0.07 per 1M tokens (7:2:1 cache hit/input/output ratio).

Prices on Baseten vary up to 32x across models, from $0.07 per 1M tokens for DeepSeek V4 Flash 0731 (max) to $2.31 per 1M tokens for Kimi K3 (low).

Yes, Baseten offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.

Yes, all 19 models on Baseten support JSON mode for structured output.

Yes, all 19 models on Baseten support function calling (tool use).

Yes, Baseten offers 15 reasoning models: Kimi K3 (max), DeepSeek V4 Pro 0813 (max), GLM-5.2 (max) (FAST), GLM-5.2 (max), DeepSeek V4 Flash 0731 (max), Kimi K3 (low), DeepSeek V4 Pro (max), Kimi K2.6, DeepSeek V4 Pro (high), Kimi K2.7 Code, Inkling, Inkling Small, GLM-4.7, gpt-oss-120b (high), and gpt-oss-120b (low). Reasoning models use extended thinking to work through complex problems before providing an answer.

Yes, all 19 models on Baseten are open weight models.

Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.

When choosing a model on Baseten, consider: intelligence (for quality-sensitive tasks), output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and features like context window size, JSON mode, or function calling support.