The Rwandan genocide is a fair contender for the top place in terms of casualties per day -- at 5-10k/day.
Another is WW2, which resulted in 70 million to 85 million casualties depending on the estimates (50-56m from the war itself, and another 19-28m from disease and famine), over the course of 2,193 days (Sep 1 1939 to Sep 2 1945) -- 32-39k/day.
A few localized bombing campaigns probably score highest on the list. For instance:
- The nuke that struck Hiroshima killed 90–146k people -- of which 70-80k killed in a mere few instants, which I'd surmise is the absolute record in history when normalized per day, and hopefully it will remain that way.
- Operation Meetinghouse (firebombing of Tokyo on Mar 9-10 1945) killed an estimated 100k people in just about 2 hours and 40 minutes -- possibly the highest rate of death when normalized per day (nearly 900k) in the history of conventional warfare.
Several WW1 battles were more localized and did excruciating amounts of casualties -- Passchendaele, Verdun, Somme... The Hundred Days Offensive, which basically ended the war, resulted in 1.8m casualties -- 18k/day.
Also worth noting are a few massacres that stood out in the medieval era and earlier. In no particular order:
The Mongols did a few wholesale massacres that captured the imaginations at the time -- Merv and Nishapur in particular. Numbers as high as a million+ are circulating on the web but I'm frankly skeptical because Persia's entire population at the time was around 2.5m. The two main sources on these seem to be Ata-Malik Juvayni and Rashid-al-Din Hamadani, neither of which were born when the events occurred.
Timur was equally murderous as he conquered Persia. Here too, the very large numbers are suspicious, albeit less so than those of the Mongols. In The History of Terrorism: From Antiquity to Al Qaeda, Chaliand and Blin (2007) put the number of deaths in Isfahan at 100-200k, while Encyclopaedia iranica and Christensen (1993; The Decline of Iranshahr) both put the number closer to 70k. There also appears to have been a witness at the time who counted more than 28 towers constructed of about 1,500 heads each -- suggestion a bit over 42k.
The Romans completely destroyed Carthage. It didn't happen overnight - the process reportedly took 7 days and 6 nights. Between 150-450k Carthagenians may have died. (The 450k number also seems high IMO: the entire Roman Empire had a population of 4-5m by around 100BC)
You'll find a few more good contenders in these two related questions:
Also, if you normalize this further so the rate of death is based on the total population of an area or whatever else expresses the death rate not in absolute but in percentage of applicable population, then by all means Napoleonic battles such as Borodino and Leipzig or some US Civil War battles like Gettysburg deserve a mention here too. As would Nazi concentration camps, which basically industrialized murder.
And if you include diseases, then think Black Death or Columbian Exchange.