vs
1:1
overall capability wins (0 ties across 2 scored). Price/speed category wins do not decide overall. Tied overall.
- Pricing3:0
- Speed2:0
- Specs1:1
Verdict
Amazon Titan Text Premier offers ~7.8x larger context (32K vs 4K); Amazon Titan Text Lite is ~4.6x cheaper on a blended token basis than Amazon Titan Text Premier.
- Amazon Titan Text Premier offers ~7.8x larger context (32K vs 4K)
- Amazon Titan Text Lite is ~4.6x cheaper on a blended token basis than Amazon Titan Text Premier
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Amazon Titan Text Lite
- $0.20
- Amazon Titan Text Premier
- $0.88
Amazon Titan Text Lite is about 4.4x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Amazon
Amazon
Family
Amazon
Amazon
License
Proprietary
Proprietary
Open weights
No
No
Release
Sep 28, 2023
Apr 30, 2024
Knowledge cutoff
-
-
API / provider
Amazon Bedrock
Amazon Bedrock
Modalities
text → text
text → text
Specs
Context window
4K
32K
Max output
4K
3K
Parameters
—
—
Pricing
Input $/1M
$0.15
$0.50
Output $/1M
$0.20
$1.50
Blended $/1M (3∶1)
$0.16
$0.75
Speed
tok/s
110
70
TTFT (s)
0.25
0.4
Reasoning
MMLU-Pro
—
—
GPQA Diamond
—
—
Humanity's Last Exam
—
—
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
—
AA-LCR v1.1
—
—
CritPt
—
—
MMMU-Pro
—
—
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
—
—
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
—
| Metric | Amazon Titan Text Lite | Amazon Titan Text Premier | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Amazon | Amazon | — |
| Family | Amazon | Amazon | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Sep 28, 2023 | Apr 30, 2024 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Amazon Bedrock | Amazon Bedrock | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 4K | 32K | B +28K |
| Max output | 4K | 3K | A +1K |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $0.15 | $0.50 | A +$0.35 |
| Output $/1M | $0.20 | $1.50 | A +$1.30 |
| Blended $/1M (3∶1) | $0.16 | $0.75 | A +$0.59 |
| Speed | |||
| tok/s | 110 | 70 | A +40 |
| TTFT (s) | 0.25 | 0.4 | A +0.15 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | — | — | — |
| Humanity's Last Exam | — | — | — |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | — | — |
| AA-LCR v1.1 | — | — | — |
| CritPt | — | — | — |
| MMMU-Pro | — | — | — |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | — | — | — |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | — | — |
Amazon Titan Text Lite
- Context
- 4K / 4K out
- Parameters
- —
- Price
- $0.15 / $0.20
- Speed
- 110 tok/s · 0.25s TTFT
- Modalities
- text → text
- License
- Proprietary
Lightweight Titan SKU (amazon.titan-text-lite-v1) on Bedrock — cheap summarization and classification before Nova Micro / Lite.
Amazon Titan Text Premier
- Context
- 32K / 3K out
- Parameters
- —
- Price
- $0.50 / $1.50
- Speed
- 70 tok/s · 0.4s TTFT
- Modalities
- text → text
- License
- Proprietary
Bedrock-native Titan flagship for enterprise RAG before the Nova family.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.
These two models have no overlapping published benchmarks in our dataset. Compare specs and pricing instead.